david/ipxe
david
/
ipxe
Archived
1
0
Fork 0
This repository has been archived on 2020-12-06. You can view files and clone it, but cannot push or open issues or pull requests.
ipxe/src/drivers/block/srp.c

829 lines
24 KiB
C
Raw Normal View History

/*
* Copyright (C) 2009 Fen Systems Ltd <mbrown@fensystems.co.uk>.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
* COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
* INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
* (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
* SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
* STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
* OF THE POSSIBILITY OF SUCH DAMAGE.
*/
FILE_LICENCE ( BSD2 );
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <ipxe/scsi.h>
#include <ipxe/xfer.h>
#include <ipxe/features.h>
#include <ipxe/srp.h>
/**
* @file
*
* SCSI RDMA Protocol
*
*/
FEATURE ( FEATURE_PROTOCOL, "SRP", DHCP_EB_FEATURE_SRP, 1 );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/** Maximum length of any initiator-to-target IU that we will send
*
* The longest IU is a SRP_CMD with no additional CDB and two direct
* data buffer descriptors, which comes to 80 bytes.
*/
#define SRP_MAX_I_T_IU_LEN 80
/* Error numbers generated by SRP login rejection */
#define EINFO_SRP_LOGIN_REJ( reason, desc ) \
__einfo_uniqify ( EINFO_EPERM, ( (reason) & 0x0f ), desc )
#define EPERM_UNKNOWN \
__einfo_error ( EINFO_EPERM_UNKNOWN )
#define EINFO_EPERM_UNKNOWN EINFO_SRP_LOGIN_REJ ( \
SRP_LOGIN_REJ_REASON_UNKNOWN, \
"Unable to establish RDMA channel, no reason specified" )
#define EPERM_INSUFFICIENT_RESOURCES \
__einfo_error ( EINFO_EPERM_INSUFFICIENT_RESOURCES )
#define EINFO_EPERM_INSUFFICIENT_RESOURCES EINFO_SRP_LOGIN_REJ ( \
SRP_LOGIN_REJ_REASON_INSUFFICIENT_RESOURCES, \
"Insufficient RDMA channel resources" )
#define EPERM_BAD_MAX_I_T_IU_LEN \
__einfo_error ( EINFO_EPERM_BAD_MAX_I_T_IU_LEN )
#define EINFO_EPERM_BAD_MAX_I_T_IU_LEN EINFO_SRP_LOGIN_REJ ( \
SRP_LOGIN_REJ_REASON_BAD_MAX_I_T_IU_LEN, \
"Requested maximum initiator to target IU length value too large" )
#define EPERM_CANNOT_ASSOCIATE \
__einfo_error ( EINFO_EPERM_CANNOT_ASSOCIATE )
#define EINFO_EPERM_CANNOT_ASSOCIATE EINFO_SRP_LOGIN_REJ ( \
SRP_LOGIN_REJ_REASON_CANNOT_ASSOCIATE, \
"Unable to associate RDMA channel with specified I_T nexus" )
#define EPERM_UNSUPPORTED_BUFFER_FORMAT \
__einfo_error ( EINFO_EPERM_UNSUPPORTED_BUFFER_FORMAT )
#define EINFO_EPERM_UNSUPPORTED_BUFFER_FORMAT EINFO_SRP_LOGIN_REJ ( \
SRP_LOGIN_REJ_REASON_UNSUPPORTED_BUFFER_FORMAT, \
"One or more requested data buffer descriptor formats not supported" )
#define EPERM_NO_MULTIPLE_CHANNELS \
__einfo_error ( EINFO_EPERM_NO_MULTIPLE_CHANNELS )
#define EINFO_EPERM_NO_MULTIPLE_CHANNELS EINFO_SRP_LOGIN_REJ ( \
SRP_LOGIN_REJ_REASON_NO_MULTIPLE_CHANNELS, \
"SRP target does not support multiple RDMA channels per I_T nexus" )
#define EPERM_NO_MORE_CHANNELS \
__einfo_error ( EINFO_EPERM_NO_MORE_CHANNELS )
#define EINFO_EPERM_NO_MORE_CHANNELS EINFO_SRP_LOGIN_REJ ( \
SRP_LOGIN_REJ_REASON_NO_MORE_CHANNELS, \
"RDMA channel limit reached for this initiator" )
#define EPERM_LOGIN_REJ( reason_nibble ) \
EUNIQ ( EINFO_EPERM, (reason_nibble), EPERM_UNKNOWN, \
EPERM_INSUFFICIENT_RESOURCES, EPERM_BAD_MAX_I_T_IU_LEN, \
EPERM_CANNOT_ASSOCIATE, EPERM_UNSUPPORTED_BUFFER_FORMAT, \
EPERM_NO_MULTIPLE_CHANNELS, EPERM_NO_MORE_CHANNELS )
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/** An SRP device */
struct srp_device {
/** Reference count */
struct refcnt refcnt;
/** SCSI command issuing interface */
struct interface scsi;
/** Underlying data transfer interface */
struct interface socket;
/** RDMA memory handle */
uint32_t memory_handle;
/** Login completed successfully */
int logged_in;
/** Initiator port ID (for boot firmware table) */
union srp_port_id initiator;
/** Target port ID (for boot firmware table) */
union srp_port_id target;
/** SCSI LUN (for boot firmware table) */
struct scsi_lun lun;
/** List of active commands */
struct list_head commands;
};
/** An SRP command */
struct srp_command {
/** Reference count */
struct refcnt refcnt;
/** SRP device */
struct srp_device *srpdev;
/** List of active commands */
struct list_head list;
/** SCSI command interface */
struct interface scsi;
/** Command tag */
uint32_t tag;
};
/**
* Get reference to SRP device
*
* @v srpdev SRP device
* @ret srpdev SRP device
*/
static inline __attribute__ (( always_inline )) struct srp_device *
srpdev_get ( struct srp_device *srpdev ) {
ref_get ( &srpdev->refcnt );
return srpdev;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/**
* Drop reference to SRP device
*
* @v srpdev SRP device
*/
static inline __attribute__ (( always_inline )) void
srpdev_put ( struct srp_device *srpdev ) {
ref_put ( &srpdev->refcnt );
}
/**
* Get reference to SRP command
*
* @v srpcmd SRP command
* @ret srpcmd SRP command
*/
static inline __attribute__ (( always_inline )) struct srp_command *
srpcmd_get ( struct srp_command *srpcmd ) {
ref_get ( &srpcmd->refcnt );
return srpcmd;
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Drop reference to SRP command
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpcmd SRP command
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static inline __attribute__ (( always_inline )) void
srpcmd_put ( struct srp_command *srpcmd ) {
ref_put ( &srpcmd->refcnt );
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Free SRP command
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v refcnt Reference count
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static void srpcmd_free ( struct refcnt *refcnt ) {
struct srp_command *srpcmd =
container_of ( refcnt, struct srp_command, refcnt );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
assert ( list_empty ( &srpcmd->list ) );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
srpdev_put ( srpcmd->srpdev );
free ( srpcmd );
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/**
* Close SRP command
*
* @v srpcmd SRP command
* @v rc Reason for close
*/
static void srpcmd_close ( struct srp_command *srpcmd, int rc ) {
struct srp_device *srpdev = srpcmd->srpdev;
if ( rc != 0 ) {
DBGC ( srpdev, "SRP %p tag %08x closed: %s\n",
srpdev, srpcmd->tag, strerror ( rc ) );
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Remove from list of commands */
if ( ! list_empty ( &srpcmd->list ) ) {
list_del ( &srpcmd->list );
INIT_LIST_HEAD ( &srpcmd->list );
srpcmd_put ( srpcmd );
}
/* Shut down interfaces */
intf_shutdown ( &srpcmd->scsi, rc );
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Close SRP device
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v rc Reason for close
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static void srpdev_close ( struct srp_device *srpdev, int rc ) {
struct srp_command *srpcmd;
struct srp_command *tmp;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( rc != 0 ) {
DBGC ( srpdev, "SRP %p closed: %s\n",
srpdev, strerror ( rc ) );
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Shut down interfaces */
intf_shutdown ( &srpdev->socket, rc );
intf_shutdown ( &srpdev->scsi, rc );
/* Shut down any active commands */
list_for_each_entry_safe ( srpcmd, tmp, &srpdev->commands, list ) {
srpcmd_get ( srpcmd );
srpcmd_close ( srpcmd, rc );
srpcmd_put ( srpcmd );
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/**
* Identify SRP command by tag
*
* @v srpdev SRP device
* @v tag Command tag
* @ret srpcmd SRP command, or NULL
*/
static struct srp_command * srp_find_tag ( struct srp_device *srpdev,
uint32_t tag ) {
struct srp_command *srpcmd;
list_for_each_entry ( srpcmd, &srpdev->commands, list ) {
if ( srpcmd->tag == tag )
return srpcmd;
}
return NULL;
}
/**
* Choose an SRP command tag
*
* @v srpdev SRP device
* @ret tag New tag, or negative error
*/
static int srp_new_tag ( struct srp_device *srpdev ) {
static uint16_t tag_idx;
unsigned int i;
for ( i = 0 ; i < 65536 ; i++ ) {
tag_idx++;
if ( srp_find_tag ( srpdev, tag_idx ) == NULL )
return tag_idx;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
return -EADDRINUSE;
}
/**
* Transmit SRP login request
*
* @v srpdev SRP device
* @v initiator Initiator port ID
* @v target Target port ID
* @v tag Command tag
* @ret rc Return status code
*/
static int srp_login ( struct srp_device *srpdev, union srp_port_id *initiator,
union srp_port_id *target, uint32_t tag ) {
struct io_buffer *iobuf;
struct srp_login_req *login_req;
int rc;
/* Allocate I/O buffer */
iobuf = xfer_alloc_iob ( &srpdev->socket, sizeof ( *login_req ) );
if ( ! iobuf )
return -ENOMEM;
/* Construct login request IU */
login_req = iob_put ( iobuf, sizeof ( *login_req ) );
memset ( login_req, 0, sizeof ( *login_req ) );
login_req->type = SRP_LOGIN_REQ;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
login_req->tag.dwords[0] = htonl ( SRP_TAG_MAGIC );
login_req->tag.dwords[1] = htonl ( tag );
login_req->max_i_t_iu_len = htonl ( SRP_MAX_I_T_IU_LEN );
login_req->required_buffer_formats = SRP_LOGIN_REQ_FMT_DDBD;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
memcpy ( &login_req->initiator, initiator,
sizeof ( login_req->initiator ) );
memcpy ( &login_req->target, target, sizeof ( login_req->target ) );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
DBGC ( srpdev, "SRP %p tag %08x LOGIN_REQ:\n", srpdev, tag );
DBGC_HDA ( srpdev, 0, iobuf->data, iob_len ( iobuf ) );
/* Send login request IU */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( ( rc = xfer_deliver_iob ( &srpdev->socket, iobuf ) ) != 0 ) {
DBGC ( srpdev, "SRP %p tag %08x could not send LOGIN_REQ: "
"%s\n", srpdev, tag, strerror ( rc ) );
return rc;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
return 0;
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Receive SRP login response
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v data SRP IU
* @v len Length of SRP IU
* @ret rc Return status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static int srp_login_rsp ( struct srp_device *srpdev,
const void *data, size_t len ) {
const struct srp_login_rsp *login_rsp = data;
/* Sanity check */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( len < sizeof ( *login_rsp ) ) {
DBGC ( srpdev, "SRP %p LOGIN_RSP too short (%zd bytes)\n",
srpdev, len );
return -EINVAL;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
DBGC ( srpdev, "SRP %p tag %08x LOGIN_RSP:\n",
srpdev, ntohl ( login_rsp->tag.dwords[1] ) );
DBGC_HDA ( srpdev, 0, data, len );
/* Mark as logged in */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
srpdev->logged_in = 1;
DBGC ( srpdev, "SRP %p logged in\n", srpdev );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Notify of window change */
xfer_window_changed ( &srpdev->scsi );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
return 0;
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Receive SRP login rejection
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v data SRP IU
* @v len Length of SRP IU
* @ret rc Return status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static int srp_login_rej ( struct srp_device *srpdev,
const void *data, size_t len ) {
const struct srp_login_rej *login_rej = data;
uint32_t reason;
/* Sanity check */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( len < sizeof ( *login_rej ) ) {
DBGC ( srpdev, "SRP %p LOGIN_REJ too short (%zd bytes)\n",
srpdev, len );
return -EINVAL;
}
reason = ntohl ( login_rej->reason );
DBGC ( srpdev, "SRP %p tag %08x LOGIN_REJ reason %08x:\n",
srpdev, ntohl ( login_rej->tag.dwords[1] ), reason );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
DBGC_HDA ( srpdev, 0, data, len );
/* Login rejection always indicates an error */
return ( SRP_LOGIN_REJ_REASON_DEFINED ( reason ) ?
-EPERM_LOGIN_REJ ( reason ) : -EACCES );
}
/**
* Transmit SRP SCSI command
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v command SCSI command
* @v tag Command tag
* @ret rc Return status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static int srp_cmd ( struct srp_device *srpdev,
struct scsi_cmd *command,
uint32_t tag ) {
struct io_buffer *iobuf;
struct srp_cmd *cmd;
struct srp_memory_descriptor *data_out;
struct srp_memory_descriptor *data_in;
int rc;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Sanity check */
if ( ! srpdev->logged_in ) {
DBGC ( srpdev, "SRP %p tag %08x cannot send CMD before "
"login completes\n", srpdev, tag );
return -EBUSY;
}
/* Allocate I/O buffer */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
iobuf = xfer_alloc_iob ( &srpdev->socket, SRP_MAX_I_T_IU_LEN );
if ( ! iobuf )
return -ENOMEM;
/* Construct base portion */
cmd = iob_put ( iobuf, sizeof ( *cmd ) );
memset ( cmd, 0, sizeof ( *cmd ) );
cmd->type = SRP_CMD;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
cmd->tag.dwords[0] = htonl ( SRP_TAG_MAGIC );
cmd->tag.dwords[1] = htonl ( tag );
memcpy ( &cmd->lun, &command->lun, sizeof ( cmd->lun ) );
memcpy ( &cmd->cdb, &command->cdb, sizeof ( cmd->cdb ) );
/* Construct data-out descriptor, if present */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( command->data_out ) {
cmd->data_buffer_formats |= SRP_CMD_DO_FMT_DIRECT;
data_out = iob_put ( iobuf, sizeof ( *data_out ) );
data_out->address =
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
cpu_to_be64 ( user_to_phys ( command->data_out, 0 ) );
data_out->handle = ntohl ( srpdev->memory_handle );
data_out->len = ntohl ( command->data_out_len );
}
/* Construct data-in descriptor, if present */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( command->data_in ) {
cmd->data_buffer_formats |= SRP_CMD_DI_FMT_DIRECT;
data_in = iob_put ( iobuf, sizeof ( *data_in ) );
data_in->address =
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
cpu_to_be64 ( user_to_phys ( command->data_in, 0 ) );
data_in->handle = ntohl ( srpdev->memory_handle );
data_in->len = ntohl ( command->data_in_len );
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
DBGC2 ( srpdev, "SRP %p tag %08x CMD " SCSI_CDB_FORMAT "\n",
srpdev, tag, SCSI_CDB_DATA ( cmd->cdb ) );
/* Send IU */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( ( rc = xfer_deliver_iob ( &srpdev->socket, iobuf ) ) != 0 ) {
DBGC ( srpdev, "SRP %p tag %08x could not send CMD: %s\n",
srpdev, tag, strerror ( rc ) );
return rc;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
return 0;
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Receive SRP SCSI response
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v data SRP IU
* @v len Length of SRP IU
* @ret rc Returns status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static int srp_rsp ( struct srp_device *srpdev,
const void *data, size_t len ) {
const struct srp_rsp *rsp = data;
struct srp_command *srpcmd;
struct scsi_rsp response;
const void *sense;
ssize_t data_out_residual_count;
ssize_t data_in_residual_count;
/* Sanity check */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( len < sizeof ( *rsp ) ) {
DBGC ( srpdev, "SRP %p RSP too short (%zd bytes)\n",
srpdev, len );
return -EINVAL;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
DBGC2 ( srpdev, "SRP %p tag %08x RSP stat %02x dores %08x dires "
"%08x valid %02x%s%s%s%s%s%s\n",
srpdev, ntohl ( rsp->tag.dwords[1] ), rsp->status,
ntohl ( rsp->data_out_residual_count ),
ntohl ( rsp->data_in_residual_count ), rsp->valid,
( ( rsp->valid & SRP_RSP_VALID_DIUNDER ) ? " diunder" : "" ),
( ( rsp->valid & SRP_RSP_VALID_DIOVER ) ? " diover" : "" ),
( ( rsp->valid & SRP_RSP_VALID_DOUNDER ) ? " dounder" : "" ),
( ( rsp->valid & SRP_RSP_VALID_DOOVER ) ? " doover" : "" ),
( ( rsp->valid & SRP_RSP_VALID_SNSVALID ) ? " sns" : "" ),
( ( rsp->valid & SRP_RSP_VALID_RSPVALID ) ? " rsp" : "" ) );
/* Identify command by tag */
srpcmd = srp_find_tag ( srpdev, ntohl ( rsp->tag.dwords[1] ) );
if ( ! srpcmd ) {
DBGC ( srpdev, "SRP %p tag %08x unrecognised RSP\n",
srpdev, ntohl ( rsp->tag.dwords[1] ) );
return -ENOENT;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Hold command reference for remainder of function */
srpcmd_get ( srpcmd );
/* Build SCSI response */
memset ( &response, 0, sizeof ( response ) );
response.status = rsp->status;
data_out_residual_count = ntohl ( rsp->data_out_residual_count );
data_in_residual_count = ntohl ( rsp->data_in_residual_count );
if ( rsp->valid & SRP_RSP_VALID_DOOVER ) {
response.overrun = data_out_residual_count;
} else if ( rsp->valid & SRP_RSP_VALID_DOUNDER ) {
response.overrun = -(data_out_residual_count);
} else if ( rsp->valid & SRP_RSP_VALID_DIOVER ) {
response.overrun = data_in_residual_count;
} else if ( rsp->valid & SRP_RSP_VALID_DIUNDER ) {
response.overrun = -(data_in_residual_count);
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
sense = srp_rsp_sense_data ( rsp );
if ( sense )
memcpy ( &response.sense, sense, sizeof ( response.sense ) );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Report SCSI response */
scsi_response ( &srpcmd->scsi, &response );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Close SCSI command */
srpcmd_close ( srpcmd, 0 );
/* Drop temporary command reference */
srpcmd_put ( srpcmd );
return 0;
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Receive SRP unrecognised response IU
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v data SRP IU
* @v len Length of SRP IU
* @ret rc Returns status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static int srp_unrecognised ( struct srp_device *srpdev,
const void *data, size_t len ) {
const struct srp_common *common = data;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
DBGC ( srpdev, "SRP %p tag %08x unrecognised IU type %02x:\n",
srpdev, ntohl ( common->tag.dwords[1] ), common->type );
DBGC_HDA ( srpdev, 0, data, len );
return -ENOTSUP;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/** SRP command SCSI interface operations */
static struct interface_operation srpcmd_scsi_op[] = {
INTF_OP ( intf_close, struct srp_command *, srpcmd_close ),
};
/** SRP command SCSI interface descriptor */
static struct interface_descriptor srpcmd_scsi_desc =
INTF_DESC ( struct srp_command, scsi, srpcmd_scsi_op );
/**
* Issue SRP SCSI command
*
* @v srpdev SRP device
* @v parent Parent interface
* @v command SCSI command
* @ret tag Command tag, or negative error
*/
static int srpdev_scsi_command ( struct srp_device *srpdev,
struct interface *parent,
struct scsi_cmd *command ) {
struct srp_command *srpcmd;
int tag;
int rc;
/* Allocate command tag */
tag = srp_new_tag ( srpdev );
if ( tag < 0 ) {
rc = tag;
goto err_tag;
}
/* Allocate and initialise structure */
srpcmd = zalloc ( sizeof ( *srpcmd ) );
if ( ! srpcmd ) {
rc = -ENOMEM;
goto err_zalloc;
}
ref_init ( &srpcmd->refcnt, srpcmd_free );
intf_init ( &srpcmd->scsi, &srpcmd_scsi_desc, &srpcmd->refcnt );
srpcmd->srpdev = srpdev_get ( srpdev );
list_add ( &srpcmd->list, &srpdev->commands );
srpcmd->tag = tag;
/* Send command IU */
if ( ( rc = srp_cmd ( srpdev, command, srpcmd->tag ) ) != 0 )
goto err_cmd;
/* Attach to parent interface, leave reference with command
* list, and return.
*/
intf_plug_plug ( &srpcmd->scsi, parent );
return srpcmd->tag;
err_cmd:
srpcmd_close ( srpcmd, rc );
err_zalloc:
err_tag:
return rc;
}
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Receive data from SRP socket
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v iobuf Datagram I/O buffer
* @v meta Data transfer metadata
* @ret rc Return status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static int srpdev_deliver ( struct srp_device *srpdev,
struct io_buffer *iobuf,
struct xfer_metadata *meta __unused ) {
struct srp_common *common = iobuf->data;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
int ( * type ) ( struct srp_device *srp, const void *data, size_t len );
int rc;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Sanity check */
if ( iob_len ( iobuf ) < sizeof ( *common ) ) {
DBGC ( srpdev, "SRP %p IU too short (%zd bytes)\n",
srpdev, iob_len ( iobuf ) );
rc = -EINVAL;
goto err;
}
/* Determine IU type */
switch ( common->type ) {
case SRP_LOGIN_RSP:
type = srp_login_rsp;
break;
case SRP_LOGIN_REJ:
type = srp_login_rej;
break;
case SRP_RSP:
type = srp_rsp;
break;
default:
type = srp_unrecognised;
break;
}
/* Handle IU */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
if ( ( rc = type ( srpdev, iobuf->data, iob_len ( iobuf ) ) ) != 0 )
goto err;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
free_iob ( iobuf );
return 0;
err:
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
DBGC ( srpdev, "SRP %p closing due to received IU (%s):\n",
srpdev, strerror ( rc ) );
DBGC_HDA ( srpdev, 0, iobuf->data, iob_len ( iobuf ) );
free_iob ( iobuf );
srpdev_close ( srpdev, rc );
return rc;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/**
* Check SRP device flow-control window
*
* @v srpdev SRP device
* @ret len Length of window
*/
static size_t srpdev_window ( struct srp_device *srpdev ) {
return ( srpdev->logged_in ? ~( ( size_t ) 0 ) : 0 );
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/**
* A (transport-independent) sBFT created by iPXE
*/
struct ipxe_sbft {
/** The table header */
struct sbft_table table;
/** The SCSI subtable */
struct sbft_scsi_subtable scsi;
/** The SRP subtable */
struct sbft_srp_subtable srp;
} __attribute__ (( packed, aligned ( 16 ) ));
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Describe SRP device in an ACPI table
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v srpdev SRP device
* @v acpi ACPI table
* @v len Length of ACPI table
* @ret rc Return status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
static int srpdev_describe ( struct srp_device *srpdev,
struct acpi_description_header *acpi,
size_t len ) {
struct ipxe_sbft *sbft =
container_of ( acpi, struct ipxe_sbft, table.acpi );
int rc;
/* Sanity check */
if ( len < sizeof ( *sbft ) )
return -ENOBUFS;
/* Populate table */
sbft->table.acpi.signature = cpu_to_le32 ( SBFT_SIG );
sbft->table.acpi.length = cpu_to_le32 ( sizeof ( *sbft ) );
sbft->table.acpi.revision = 1;
sbft->table.scsi_offset =
cpu_to_le16 ( offsetof ( typeof ( *sbft ), scsi ) );
memcpy ( &sbft->scsi.lun, &srpdev->lun, sizeof ( sbft->scsi.lun ) );
sbft->table.srp_offset =
cpu_to_le16 ( offsetof ( typeof ( *sbft ), srp ) );
memcpy ( &sbft->srp.initiator, &srpdev->initiator,
sizeof ( sbft->srp.initiator ) );
memcpy ( &sbft->srp.target, &srpdev->target,
sizeof ( sbft->srp.target ) );
/* Ask transport layer to describe transport-specific portions */
if ( ( rc = acpi_describe ( &srpdev->socket, acpi, len ) ) != 0 ) {
DBGC ( srpdev, "SRP %p cannot describe transport layer: %s\n",
srpdev, strerror ( rc ) );
return rc;
}
return 0;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/** SRP device socket interface operations */
static struct interface_operation srpdev_socket_op[] = {
INTF_OP ( xfer_deliver, struct srp_device *, srpdev_deliver ),
INTF_OP ( intf_close, struct srp_device *, srpdev_close ),
};
/** SRP device socket interface descriptor */
static struct interface_descriptor srpdev_socket_desc =
INTF_DESC_PASSTHRU ( struct srp_device, socket, srpdev_socket_op,
scsi );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/** SRP device SCSI interface operations */
static struct interface_operation srpdev_scsi_op[] = {
INTF_OP ( scsi_command, struct srp_device *, srpdev_scsi_command ),
INTF_OP ( xfer_window, struct srp_device *, srpdev_window ),
INTF_OP ( intf_close, struct srp_device *, srpdev_close ),
INTF_OP ( acpi_describe, struct srp_device *, srpdev_describe ),
};
/** SRP device SCSI interface descriptor */
static struct interface_descriptor srpdev_scsi_desc =
INTF_DESC_PASSTHRU ( struct srp_device, scsi, srpdev_scsi_op, socket );
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/**
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* Open SRP device
*
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
* @v block Block control interface
* @v socket Socket interface
* @v initiator Initiator port ID
* @v target Target port ID
* @v memory_handle RDMA memory handle
* @v lun SCSI LUN
* @ret rc Return status code
*/
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
int srp_open ( struct interface *block, struct interface *socket,
union srp_port_id *initiator, union srp_port_id *target,
uint32_t memory_handle, struct scsi_lun *lun ) {
struct srp_device *srpdev;
int tag;
int rc;
/* Allocate and initialise structure */
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
srpdev = zalloc ( sizeof ( *srpdev ) );
if ( ! srpdev ) {
rc = -ENOMEM;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
goto err_zalloc;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
ref_init ( &srpdev->refcnt, NULL );
intf_init ( &srpdev->scsi, &srpdev_scsi_desc, &srpdev->refcnt );
intf_init ( &srpdev->socket, &srpdev_socket_desc, &srpdev->refcnt );
INIT_LIST_HEAD ( &srpdev->commands );
srpdev->memory_handle = memory_handle;
DBGC ( srpdev, "SRP %p %08x%08x%08x%08x->%08x%08x%08x%08x\n", srpdev,
ntohl ( initiator->dwords[0] ), ntohl ( initiator->dwords[1] ),
ntohl ( initiator->dwords[2] ), ntohl ( initiator->dwords[3] ),
ntohl ( target->dwords[0] ), ntohl ( target->dwords[1] ),
ntohl ( target->dwords[2] ), ntohl ( target->dwords[3] ) );
/* Preserve parameters required for boot firmware table */
memcpy ( &srpdev->initiator, initiator, sizeof ( srpdev->initiator ) );
memcpy ( &srpdev->target, target, sizeof ( srpdev->target ) );
memcpy ( &srpdev->lun, lun, sizeof ( srpdev->lun ) );
/* Attach to socket interface and initiate login */
intf_plug_plug ( &srpdev->socket, socket );
tag = srp_new_tag ( srpdev );
assert ( tag >= 0 ); /* Cannot fail when no commands in progress */
if ( ( rc = srp_login ( srpdev, initiator, target, tag ) ) != 0 )
goto err_login;
/* Attach SCSI device to parent interface */
if ( ( rc = scsi_open ( block, &srpdev->scsi, lun ) ) != 0 ) {
DBGC ( srpdev, "SRP %p could not create SCSI device: %s\n",
srpdev, strerror ( rc ) );
goto err_scsi_open;
}
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
/* Mortalise self and return */
ref_put ( &srpdev->refcnt );
return 0;
[block] Replace gPXE block-device API with an iPXE asynchronous interface The block device interface used in gPXE predates the invention of even the old gPXE data-transfer interface, let alone the current iPXE generic asynchronous interface mechanism. Bring this old code up to date, with the following benefits: o Block device commands can be cancelled by the requestor. The INT 13 layer uses this to provide a global timeout on all INT 13 calls, with the result that an unexpected passive failure mode (such as an iSCSI target ACKing the request but never sending a response) will lead to a timeout that gets reported back to the INT 13 user, rather than simply freezing the system. o INT 13,00 (reset drive) is now able to reset the underlying block device. INT 13 users, such as DOS, that use INT 13,00 as a method for error recovery now have a chance of recovering. o All block device commands are tagged, with a numerical tag that will show up in debugging output and in packet captures; this will allow easier interpretation of bug reports that include both sources of information. o The extremely ugly hacks used to generate the boot firmware tables have been eradicated and replaced with a generic acpi_describe() method (exploiting the ability of iPXE interfaces to pass through methods to an underlying interface). The ACPI tables are now built in a shared data block within .bss16, rather than each requiring dedicated space in .data16. o The architecture-independent concept of a SAN device has been exposed to the iPXE core through the sanboot API, which provides calls to hook, unhook, boot, and describe SAN devices. This allows for much more flexible usage patterns (such as hooking an empty SAN device and then running an OS installer via TFTP). Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-09-03 17:11:51 +02:00
err_scsi_open:
err_login:
srpdev_close ( srpdev, rc );
ref_put ( &srpdev->refcnt );
err_zalloc:
return rc;
}