david/ipxe
david
/
ipxe
Archived
1
0
Fork 0
Code
This repository has been archived on 2020-12-06.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/ibft.c

688 lines
19 KiB
C
Raw Normal View History

Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/*
* Copyright Fen Systems Ltd. 2007. Portions of this code are derived
* from IBM Corporation Sample Programs. Copyright IBM Corporation
* 2004, 2007. All rights reserved.
*
* Permission is hereby granted, free of charge, to any person
* obtaining a copy of this software and associated documentation
* files (the "Software"), to deal in the Software without
* restriction, including without limitation the rights to use, copy,
* modify, merge, publish, distribute, sublicense, and/or sell copies
* of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be
* included in all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*
*/
[legal] Add a selection of FILE_LICENCE declarations Add FILE_LICENCE declarations to almost all files that make up the various standard builds of gPXE.
2009-05-01 16:41:06 +02:00
FILE_LICENCE ( BSD2 );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
#include <stdint.h>
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
#include <stdlib.h>
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
#include <stdio.h>
#include <string.h>
#include <errno.h>
#include <byteswap.h>
[build] Rename gPXE to iPXE Access to the gpxe.org and etherboot.org domains and associated resources has been revoked by the registrant of the domain. Work around this problem by renaming project from gPXE to iPXE, and updating URLs to match. Also update README, LOG and COPYRIGHTS to remove obsolete information. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-04-19 21:16:01 +02:00
#include <ipxe/pci.h>
#include <ipxe/acpi.h>
#include <ipxe/in.h>
#include <ipxe/netdevice.h>
#include <ipxe/ethernet.h>
[iscsi] Include 802.1Q VLAN identifier in iBFT The iBFT NIC section has a VLAN field which must be filled in so that iSCSI booting works over VLANs. Unfortunately it is unclear from the IBM specification linked in ibft.c whether the VLAN field is just the 802.1Q VLAN Identifier or the full 802.1Q TCI. For now just fill in the VID, the Priority Code Point and Drop Eligible Indicator could be set in the future if it turns out they should be present too. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Modified-by: Michael Brown <mcb30@ipxe.org> Signed-off-by: Michael Brown <mcb30@ipxe.org>
2013-03-01 16:32:16 +01:00
#include <ipxe/vlan.h>
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
#include <ipxe/tcpip.h>
[build] Rename gPXE to iPXE Access to the gpxe.org and etherboot.org domains and associated resources has been revoked by the registrant of the domain. Work around this problem by renaming project from gPXE to iPXE, and updating URLs to match. Also update README, LOG and COPYRIGHTS to remove obsolete information. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-04-19 21:16:01 +02:00
#include <ipxe/dhcp.h>
#include <ipxe/iscsi.h>
#include <ipxe/ibft.h>
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/** @file
*
* iSCSI boot firmware table
*
* The information in this file is derived from the document "iSCSI
* Boot Firmware Table (iBFT)" as published by IBM at
*
* ftp://ftp.software.ibm.com/systems/support/system_x_pdf/ibm_iscsi_boot_firmware_table_v1.02.pdf
*
*/
[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] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
* iSCSI string 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
*
* This is an internal structure that we use to keep track of the
* allocation of string data.
*/
struct ibft_strings {
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/** Strings data */
char *data;
/** Starting offset of strings */
size_t start;
/** Total length */
[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
size_t len;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
};
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/**
* Align structure within iBFT
*
* @v len Unaligned length (or offset)
* @ret len Aligned length (or offset)
*/
static inline size_t ibft_align ( size_t len ) {
return ( ( len + IBFT_ALIGN - 1 ) & ~( IBFT_ALIGN - 1 ) );
}
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/**
* Fill in an IP address field within iBFT
*
* @v ipaddr IP address field
* @v in IPv4 address
*/
static void ibft_set_ipaddr ( struct ibft_ipaddr *ipaddr, struct in_addr in ) {
[iscsi] Fix minor typo in ibft_set_ipaddr() Signed-off-by: Michael Brown <mcb30@ipxe.org>
2011-02-25 00:39:38 +01:00
memset ( ipaddr, 0, sizeof ( *ipaddr ) );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
if ( in.s_addr ) {
ipaddr->in = in;
ipaddr->ones = 0xffff;
}
}
/**
[Settings] Remove assumption that all settings have DHCP tag values Allow for settings to be described by something other than a DHCP option tag if desirable. Currently used only for the MAC address setting. Separate out fake DHCP packet creation code from dhcp.c to fakedhcp.c. Remove notion of settings from dhcppkt.c. Rationalise dhcp.c to use settings API only for final registration of the DHCP options, rather than using {store,fetch}_setting throughout.
2008-03-25 21:46:16 +01:00
* Fill in an IP address within iBFT from configuration setting
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
*
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
* @v settings Parent settings block, or NULL
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
* @v ipaddr IP address field
[Settings] Remove assumption that all settings have DHCP tag values Allow for settings to be described by something other than a DHCP option tag if desirable. Currently used only for the MAC address setting. Separate out fake DHCP packet creation code from dhcp.c to fakedhcp.c. Remove notion of settings from dhcppkt.c. Rationalise dhcp.c to use settings API only for final registration of the DHCP options, rather than using {store,fetch}_setting throughout.
2008-03-25 21:46:16 +01:00
* @v setting Configuration setting
[iscsi] Include both DNS addresses in iBFT, if available Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-12-09 16:35:02 +01:00
* @v count Maximum number of IP addresses
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
*/
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
static void ibft_set_ipaddr_setting ( struct settings *settings,
struct ibft_ipaddr *ipaddr,
[settings] Explicitly separate the concept of a completed fetched setting The fetch_setting() family of functions may currently modify the definition of the specified setting (e.g. to add missing type information). Clean up this interface by requiring callers to provide an explicit buffer to contain the completed definition of the fetched setting, if required. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2013-12-03 17:48:56 +01:00
const struct setting *setting,
[iscsi] Include both DNS addresses in iBFT, if available Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-12-09 16:35:02 +01:00
unsigned int count ) {
struct in_addr in[count];
unsigned int i;
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
fetch_ipv4_array_setting ( settings, setting, in, count );
[iscsi] Include both DNS addresses in iBFT, if available Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-12-09 16:35:02 +01:00
for ( i = 0 ; i < count ; i++ ) {
ibft_set_ipaddr ( &ipaddr[i], in[i] );
}
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
}
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
/**
* Read IP address from iBFT (for debugging)
*
* @v strings iBFT string block descriptor
* @v string String field
* @ret ipaddr IP address string
*/
static const char * ibft_ipaddr ( struct ibft_ipaddr *ipaddr ) {
return inet_ntoa ( ipaddr->in );
}
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/**
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
* Allocate a string within iBFT
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
*
* @v strings iBFT string block descriptor
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
* @v string String field to fill in
* @v len Length of string to allocate (excluding NUL)
[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
* @ret dest String destination, or NULL
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +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
static char * ibft_alloc_string ( struct ibft_strings *strings,
struct ibft_string *string, size_t len ) {
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
size_t new_len;
char *new_data;
char *dest;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/* Extend string data buffer */
new_len = ( strings->len + len + 1 /* NUL */ );
new_data = realloc ( strings->data, new_len );
if ( ! new_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
return NULL;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
strings->data = new_data;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/* Fill in string field */
string->offset = cpu_to_le16 ( strings->start + strings->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
string->len = cpu_to_le16 ( len );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/* Zero string */
dest = ( strings->data + strings->len );
memset ( dest, 0, ( len + 1 /* NUL */ ) );
/* Update allocated length */
strings->len = new_len;
return dest;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
}
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
/**
* Fill in a string field within iBFT
*
* @v strings iBFT string block descriptor
* @v string String field
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
* @v data String to fill in, or NULL
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
* @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 ibft_set_string ( struct ibft_strings *strings,
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
struct ibft_string *string, const char *data ) {
char *dest;
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
if ( ! data )
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
dest = ibft_alloc_string ( strings, string, strlen ( data ) );
if ( ! dest )
return -ENOBUFS;
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
strcpy ( dest, data );
return 0;
}
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/**
[Settings] Remove assumption that all settings have DHCP tag values Allow for settings to be described by something other than a DHCP option tag if desirable. Currently used only for the MAC address setting. Separate out fake DHCP packet creation code from dhcp.c to fakedhcp.c. Remove notion of settings from dhcppkt.c. Rationalise dhcp.c to use settings API only for final registration of the DHCP options, rather than using {store,fetch}_setting throughout.
2008-03-25 21:46:16 +01:00
* Fill in a string field within iBFT from configuration setting
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
*
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
* @v settings Parent settings block, or NULL
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
* @v strings iBFT string block descriptor
* @v string String field
[Settings] Remove assumption that all settings have DHCP tag values Allow for settings to be described by something other than a DHCP option tag if desirable. Currently used only for the MAC address setting. Separate out fake DHCP packet creation code from dhcp.c to fakedhcp.c. Remove notion of settings from dhcppkt.c. Rationalise dhcp.c to use settings API only for final registration of the DHCP options, rather than using {store,fetch}_setting throughout.
2008-03-25 21:46:16 +01:00
* @v setting Configuration setting
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
* @ret rc Return status code
*/
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
static int ibft_set_string_setting ( struct settings *settings,
struct ibft_strings *strings,
[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
struct ibft_string *string,
[settings] Explicitly separate the concept of a completed fetched setting The fetch_setting() family of functions may currently modify the definition of the specified setting (e.g. to add missing type information). Clean up this interface by requiring callers to provide an explicit buffer to contain the completed definition of the fetched setting, if required. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2013-12-03 17:48:56 +01:00
const struct setting *setting ) {
struct settings *origin;
struct setting fetched;
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
int len;
char *dest;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
len = fetch_setting ( settings, setting, &origin, &fetched, NULL, 0 );
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
if ( len < 0 ) {
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
string->offset = 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
string->len = 0;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +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
dest = ibft_alloc_string ( strings, string, len );
if ( ! dest )
return -ENOBUFS;
[settings] Explicitly separate the concept of a completed fetched setting The fetch_setting() family of functions may currently modify the definition of the specified setting (e.g. to add missing type information). Clean up this interface by requiring callers to provide an explicit buffer to contain the completed definition of the fetched setting, if required. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2013-12-03 17:48:56 +01:00
fetch_string_setting ( origin, &fetched, dest, ( len + 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
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
return 0;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
}
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
/**
* Read string from iBFT (for debugging)
*
* @v strings iBFT string block descriptor
* @v string String field
* @ret data String content (or "<empty>")
*/
[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 const char * ibft_string ( struct ibft_strings *strings,
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
struct ibft_string *string ) {
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
size_t offset = le16_to_cpu ( string->offset );
return ( offset ? ( strings->data + offset - strings->start ) : NULL );
}
/**
* Check if network device is required for the iBFT
*
* @v netdev Network device
* @ret is_required Network device is required
*/
static int ibft_netdev_is_required ( struct net_device *netdev ) {
struct iscsi_session *iscsi;
struct sockaddr_tcpip *st_target;
list_for_each_entry ( iscsi, &ibft_model.descs, desc.list ) {
st_target = ( struct sockaddr_tcpip * ) &iscsi->target_sockaddr;
if ( tcpip_netdev ( st_target ) == netdev )
return 1;
}
return 0;
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
}
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/**
* Fill in NIC portion of iBFT
*
* @v nic NIC portion of iBFT
* @v strings iBFT string block descriptor
* @v netdev Network device
* @ret rc Return status code
*/
static int ibft_fill_nic ( struct ibft_nic *nic,
[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
struct ibft_strings *strings,
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
struct net_device *netdev ) {
[iscsi] Use the "Ethernet-compatible" MAC address in the iBFT
2009-10-23 23:18:04 +02:00
struct ll_protocol *ll_protocol = netdev->ll_protocol;
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
struct in_addr netmask_addr = { 0 };
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
unsigned int netmask_count = 0;
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
struct settings *parent = netdev_settings ( netdev );
[iscsi] Include IP address origin in iBFT The iBFT includes an "origin" field to indicate the source of the IP address. We use the heuristic of assuming that the source should be "manual" if the IP address originates directly from the network device settings block, and "DHCP" otherwise. This is an imperfect guess, but is likely to be correct in most common situations. Originally-implemented-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:05:48 +02:00
struct settings *origin;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
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
/* Fill in common header */
nic->header.structure_id = IBFT_STRUCTURE_ID_NIC;
nic->header.version = 1;
nic->header.length = cpu_to_le16 ( sizeof ( *nic ) );
nic->header.flags = ( IBFT_FL_NIC_BLOCK_VALID |
IBFT_FL_NIC_FIRMWARE_BOOT_SELECTED );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d is %s\n", nic->header.index, netdev->name );
[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
[iscsi] Include IP address origin in iBFT The iBFT includes an "origin" field to indicate the source of the IP address. We use the heuristic of assuming that the source should be "manual" if the IP address originates directly from the network device settings block, and "DHCP" otherwise. This is an imperfect guess, but is likely to be correct in most common situations. Originally-implemented-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:05:48 +02:00
/* Determine origin of IP address */
fetch_setting ( parent, &ip_setting, &origin, NULL, NULL, 0 );
nic->origin = ( ( origin == parent ) ?
IBFT_NIC_ORIGIN_MANUAL : IBFT_NIC_ORIGIN_DHCP );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d origin = %d\n", nic->header.index, nic->origin );
[iscsi] Include IP address origin in iBFT The iBFT includes an "origin" field to indicate the source of the IP address. We use the heuristic of assuming that the source should be "manual" if the IP address originates directly from the network device settings block, and "DHCP" otherwise. This is an imperfect guess, but is likely to be correct in most common situations. Originally-implemented-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:05:48 +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
/* Extract values from configuration settings */
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
ibft_set_ipaddr_setting ( parent, &nic->ip_address, &ip_setting, 1 );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d IP = %s\n",
nic->header.index, ibft_ipaddr ( &nic->ip_address ) );
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
ibft_set_ipaddr_setting ( parent, &nic->gateway, &gateway_setting, 1 );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d gateway = %s\n",
nic->header.index, ibft_ipaddr ( &nic->gateway ) );
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
ibft_set_ipaddr_setting ( NULL, &nic->dns[0], &dns_setting,
[iscsi] Include both DNS addresses in iBFT, if available Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-12-09 16:35:02 +01:00
( sizeof ( nic->dns ) /
sizeof ( nic->dns[0] ) ) );
[iscsi] Include DHCP server address in iBFT Signed-off-by: Michael Brown <mcb30@ipxe.org>
2016-03-29 20:38:18 +02:00
ibft_set_ipaddr_setting ( parent, &nic->dhcp, &dhcp_server_setting, 1 );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d DNS = %s",
nic->header.index, ibft_ipaddr ( &nic->dns[0] ) );
[iscsi] Include both DNS addresses in iBFT, if available Signed-off-by: Michael Brown <mcb30@ipxe.org>
2010-12-09 16:35:02 +01:00
DBG ( ", %s\n", ibft_ipaddr ( &nic->dns[1] ) );
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
if ( ( rc = ibft_set_string_setting ( NULL, strings, &nic->hostname,
[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
&hostname_setting ) ) != 0 )
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
return rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d hostname = %s\n",
nic->header.index, ibft_string ( strings, &nic->hostname ) );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/* Derive subnet mask prefix from subnet mask */
[iscsi] Read IPv4 settings only from the relevant network device Signed-off-by: Michael Brown <mcb30@ipxe.org>
2014-06-12 17:08:28 +02:00
fetch_ipv4_setting ( parent, &netmask_setting, &netmask_addr );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
while ( netmask_addr.s_addr ) {
if ( netmask_addr.s_addr & 0x1 )
netmask_count++;
netmask_addr.s_addr >>= 1;
}
nic->subnet_mask_prefix = netmask_count;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d subnet = /%d\n",
nic->header.index, nic->subnet_mask_prefix );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/* Extract values from net-device configuration */
[iscsi] Include 802.1Q VLAN identifier in iBFT The iBFT NIC section has a VLAN field which must be filled in so that iSCSI booting works over VLANs. Unfortunately it is unclear from the IBM specification linked in ibft.c whether the VLAN field is just the 802.1Q VLAN Identifier or the full 802.1Q TCI. For now just fill in the VID, the Priority Code Point and Drop Eligible Indicator could be set in the future if it turns out they should be present too. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Modified-by: Michael Brown <mcb30@ipxe.org> Signed-off-by: Michael Brown <mcb30@ipxe.org>
2013-03-01 16:32:16 +01:00
nic->vlan = cpu_to_le16 ( vlan_tag ( netdev ) );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d VLAN = %02x\n",
nic->header.index, le16_to_cpu ( nic->vlan ) );
[iscsi] Use the "Ethernet-compatible" MAC address in the iBFT
2009-10-23 23:18:04 +02:00
if ( ( rc = ll_protocol->eth_addr ( netdev->ll_addr,
nic->mac_address ) ) != 0 ) {
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "Could not determine %s MAC: %s\n",
netdev->name, strerror ( rc ) );
[iscsi] Use the "Ethernet-compatible" MAC address in the iBFT
2009-10-23 23:18:04 +02:00
return rc;
}
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d MAC = %s\n",
nic->header.index, eth_ntoa ( nic->mac_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
nic->pci_bus_dev_func = cpu_to_le16 ( netdev->dev->desc.location );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT NIC %d PCI = %04x\n",
nic->header.index, le16_to_cpu ( nic->pci_bus_dev_func ) );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
return 0;
}
/**
* Fill in Initiator portion of iBFT
*
* @v initiator Initiator portion of iBFT
* @v strings iBFT string block descriptor
* @ret rc Return status code
*/
static int ibft_fill_initiator ( struct ibft_initiator *initiator,
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
struct ibft_strings *strings ) {
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
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
/* Fill in common header */
initiator->header.structure_id = IBFT_STRUCTURE_ID_INITIATOR;
initiator->header.version = 1;
initiator->header.length = cpu_to_le16 ( sizeof ( *initiator ) );
initiator->header.flags = ( IBFT_FL_INITIATOR_BLOCK_VALID |
IBFT_FL_INITIATOR_FIRMWARE_BOOT_SELECTED );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/* Fill in initiator name */
if ( ( rc = ibft_set_string_setting ( NULL, strings,
&initiator->initiator_name,
&initiator_iqn_setting ) ) != 0 )
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
return rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT initiator name = %s\n",
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
ibft_string ( strings, &initiator->initiator_name ) );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
return 0;
}
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/**
* Fill in Target NIC association
*
* @v target Target portion of iBFT
* @v iscsi iSCSI session
* @ret rc Return status code
*/
static int ibft_fill_target_nic_association ( struct ibft_target *target,
struct iscsi_session *iscsi ) {
struct sockaddr_tcpip *st_target =
( struct sockaddr_tcpip * ) &iscsi->target_sockaddr;
struct net_device *associated;
struct net_device *netdev;
/* Find network device used to reach target */
associated = tcpip_netdev ( st_target );
if ( ! associated ) {
DBG ( "iBFT target %d has no net device\n",
target->header.index );
return -EHOSTUNREACH;
}
/* Calculate association */
for_each_netdev ( netdev ) {
if ( netdev == associated ) {
DBG ( "iBFT target %d uses NIC %d (%s)\n",
target->header.index, target->nic_association,
netdev->name );
return 0;
}
if ( ! ibft_netdev_is_required ( netdev ) )
continue;
target->nic_association++;
}
DBG ( "iBFT target %d has impossible NIC %s\n",
target->header.index, netdev->name );
return -EINVAL;
}
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
/**
* Fill in Target CHAP portion of iBFT
*
* @v target Target portion of iBFT
* @v strings iBFT string block descriptor
* @v iscsi iSCSI session
* @ret rc Return status code
*/
static int ibft_fill_target_chap ( struct ibft_target *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
struct ibft_strings *strings,
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
struct iscsi_session *iscsi ) {
int rc;
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
if ( ! ( iscsi->status & ISCSI_STATUS_AUTH_FORWARD_REQUIRED ) )
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
return 0;
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
assert ( iscsi->initiator_username );
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
assert ( iscsi->initiator_password );
target->chap_type = IBFT_CHAP_ONE_WAY;
if ( ( rc = ibft_set_string ( strings, &target->chap_name,
iscsi->initiator_username ) ) != 0 )
return rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT target %d username = %s\n", target->header.index,
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
ibft_string ( strings, &target->chap_name ) );
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
if ( ( rc = ibft_set_string ( strings, &target->chap_secret,
iscsi->initiator_password ) ) != 0 )
return rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT target %d password = <redacted>\n", target->header.index );
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
return 0;
}
/**
* Fill in Target Reverse CHAP portion of iBFT
*
* @v target Target portion of iBFT
* @v strings iBFT string block descriptor
* @v iscsi iSCSI session
* @ret rc Return status code
*/
static int ibft_fill_target_reverse_chap ( struct ibft_target *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
struct ibft_strings *strings,
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
struct iscsi_session *iscsi ) {
int rc;
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
if ( ! ( iscsi->status & ISCSI_STATUS_AUTH_REVERSE_REQUIRED ) )
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
return 0;
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
assert ( iscsi->initiator_username );
assert ( iscsi->initiator_password );
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
assert ( iscsi->target_username );
assert ( iscsi->target_password );
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
target->chap_type = IBFT_CHAP_MUTUAL;
if ( ( rc = ibft_set_string ( strings, &target->reverse_chap_name,
iscsi->target_username ) ) != 0 )
return rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT target %d reverse username = %s\n", target->header.index,
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
ibft_string ( strings, &target->chap_name ) );
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
if ( ( rc = ibft_set_string ( strings, &target->reverse_chap_secret,
iscsi->target_password ) ) != 0 )
return rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT target %d reverse password = <redacted>\n",
target->header.index );
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
return 0;
}
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/**
* Fill in Target portion of iBFT
*
* @v target Target portion of iBFT
* @v strings iBFT string block descriptor
* @v iscsi iSCSI session
* @ret rc Return status code
*/
static int ibft_fill_target ( struct ibft_target *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
struct ibft_strings *strings,
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
struct iscsi_session *iscsi ) {
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
struct sockaddr_tcpip *st_target =
( struct sockaddr_tcpip * ) &iscsi->target_sockaddr;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
struct sockaddr_in *sin_target =
( struct sockaddr_in * ) &iscsi->target_sockaddr;
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
/* Fill in common header */
target->header.structure_id = IBFT_STRUCTURE_ID_TARGET;
target->header.version = 1;
target->header.length = cpu_to_le16 ( sizeof ( *target ) );
target->header.flags = ( IBFT_FL_TARGET_BLOCK_VALID |
IBFT_FL_TARGET_FIRMWARE_BOOT_SELECTED );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
/* Fill in Target values */
ibft_set_ipaddr ( &target->ip_address, sin_target->sin_addr );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT target %d IP = %s\n",
target->header.index, ibft_ipaddr ( &target->ip_address ) );
target->socket = cpu_to_le16 ( ntohs ( st_target->st_port ) );
DBG ( "iBFT target %d port = %d\n",
target->header.index, target->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
memcpy ( &target->boot_lun, &iscsi->lun, sizeof ( target->boot_lun ) );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT target %d boot LUN = " SCSI_LUN_FORMAT "\n",
target->header.index, SCSI_LUN_DATA ( target->boot_lun ) );
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
if ( ( rc = ibft_set_string ( strings, &target->target_name,
[Settings] Use a settings applicator in ibft.c
2008-03-21 01:49:44 +01:00
iscsi->target_iqn ) ) != 0 )
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
return rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
DBG ( "iBFT target %d name = %s\n", target->header.index,
[iscsi] Include credentials in iBFT only if used during iSCSI login Avoid passing credentials in the iBFT that were available but not required for login. This works around a problem in the Microsoft iSCSI initiator, which will refuse to initiate sessions if the CHAP password is fewer than 12 characters, even if the target ends up not asking for CHAP authentication.
2009-02-20 22:41:00 +01:00
ibft_string ( strings, &target->target_name ) );
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
if ( ( rc = ibft_fill_target_nic_association ( target, iscsi ) ) != 0 )
return rc;
[iSCSI] Add support for mutual CHAP Allow initiator to verify target authentication using CHAP.
2008-08-11 04:12:38 +02:00
if ( ( rc = ibft_fill_target_chap ( target, strings, iscsi ) ) != 0 )
return rc;
if ( ( rc = ibft_fill_target_reverse_chap ( target, strings,
iscsi ) ) != 0 )
return rc;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
return 0;
}
/**
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
* Check if iBFT descriptor is complete
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
*
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
* @v desc ACPI descriptor
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
* @ret rc Return status code
*/
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
static int ibft_complete ( struct acpi_descriptor *desc ) {
struct iscsi_session *iscsi =
container_of ( desc, struct iscsi_session, desc );
/* Fail if we do not yet have the target address */
if ( ! iscsi->target_sockaddr.sa_family )
return -EAGAIN;
return 0;
}
/**
* Install iBFT
*
* @v install Installation method
* @ret rc Return status code
*/
static int ibft_install ( int ( * install ) ( struct acpi_header *acpi ) ) {
[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
struct net_device *netdev;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
struct iscsi_session *iscsi;
struct ibft_table *table;
struct ibft_initiator *initiator;
struct ibft_nic *nic;
struct ibft_target *target;
struct ibft_strings strings;
struct acpi_header *acpi;
void *data;
unsigned int targets = 0;
unsigned int pairs = 0;
size_t offset = 0;
size_t table_len;
size_t control_len;
size_t initiator_offset;
size_t nic_offset;
size_t target_offset;
size_t strings_offset;
size_t len;
unsigned int i;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
int rc;
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/* Calculate table sizes and offsets */
list_for_each_entry ( iscsi, &ibft_model.descs, desc.list )
targets++;
pairs = ( sizeof ( table->control.pair ) /
sizeof ( table->control.pair[0] ) );
if ( pairs < targets )
pairs = targets;
offset = offsetof ( typeof ( *table ), control.pair );
offset += ( pairs * sizeof ( table->control.pair[0] ) );
table_len = offset;
control_len = ( table_len - offsetof ( typeof ( *table ), control ) );
offset = ibft_align ( offset );
initiator_offset = offset;
offset += ibft_align ( sizeof ( *initiator ) );
nic_offset = offset;
offset += ( pairs * ibft_align ( sizeof ( *nic ) ) );
target_offset = offset;
offset += ( pairs * ibft_align ( sizeof ( *target ) ) );
strings_offset = offset;
strings.data = NULL;
strings.start = strings_offset;
strings.len = 0;
len = offset;
[iscsi] Do not install iBFT when no iSCSI targets exist Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-28 19:45:23 +02:00
/* Do nothing if no targets exist */
if ( ! targets ) {
rc = 0;
goto no_targets;
}
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/* Allocate table */
data = zalloc ( len );
if ( ! data ) {
rc = -ENOMEM;
goto err_alloc;
[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
}
/* Fill in Control block */
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
table = data;
table->control.header.structure_id = IBFT_STRUCTURE_ID_CONTROL;
table->control.header.version = 1;
table->control.header.length = cpu_to_le16 ( control_len );
/* Fill in Initiator block */
initiator = ( data + initiator_offset );
table->control.initiator = cpu_to_le16 ( initiator_offset );
if ( ( rc = ibft_fill_initiator ( initiator, &strings ) ) != 0 )
goto err_initiator;
/* Fill in NIC blocks */
i = 0;
for_each_netdev ( netdev ) {
if ( ! ibft_netdev_is_required ( netdev ) )
continue;
assert ( i < pairs );
table->control.pair[i].nic = nic_offset;
nic = ( data + nic_offset );
nic->header.index = i;
if ( ( rc = ibft_fill_nic ( nic, &strings, netdev ) ) != 0 )
goto err_nic;
i++;
nic_offset += ibft_align ( sizeof ( *nic ) );
}
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/* Fill in Target blocks */
i = 0;
list_for_each_entry ( iscsi, &ibft_model.descs, desc.list ) {
assert ( i < pairs );
table->control.pair[i].target = target_offset;
target = ( data + target_offset );
target->header.index = i;
if ( ( rc = ibft_fill_target ( target, &strings, iscsi ) ) != 0)
goto err_target;
i++;
target_offset += ibft_align ( sizeof ( *target ) );
}
/* Reallocate table to include space for strings */
len += strings.len;
acpi = realloc ( data, len );
if ( ! acpi )
goto err_realloc;
data = NULL;
/* Fill in ACPI header */
acpi->signature = cpu_to_le32 ( IBFT_SIG );
acpi->length = cpu_to_le32 ( len );
acpi->revision = 1;
/* Append strings */
memcpy ( ( ( ( void * ) acpi ) + strings_offset ), strings.data,
strings.len );
/* Install ACPI table */
if ( ( rc = install ( acpi ) ) != 0 ) {
DBG ( "iBFT could not install: %s\n", strerror ( rc ) );
goto err_install;
}
err_install:
free ( acpi );
err_realloc:
err_target:
err_nic:
err_initiator:
free ( data );
err_alloc:
[iscsi] Do not install iBFT when no iSCSI targets exist Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-28 19:45:23 +02:00
no_targets:
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
free ( strings.data );
return rc;
Add iBFT code derived from IBM document. No Microsoft intellectual property was harmed in the production of this code.
2007-08-10 02:44:20 +02:00
}
[block] Describe all SAN devices via ACPI tables Describe all SAN devices via ACPI tables such as the iBFT. For tables that can describe only a single device (i.e. the aBFT and sBFT), one table is installed per device. For multi-device tables (i.e. the iBFT), all devices are described in a single table. An underlying SAN device connection may be closed at the time that we need to construct an ACPI table. We therefore introduce the concept of an "ACPI descriptor" which enables the SAN boot code to maintain an opaque pointer to the underlying object, and an "ACPI model" which can build tables from a list of such descriptors. This separates the lifecycles of ACPI descriptions from the lifecycles of the block device interfaces, and allows for construction of the ACPI tables even if the block device interface has been closed. For a multipath SAN device, iPXE will wait until sufficient information is available to describe all devices but will not wait for all paths to connect successfully. For example: with a multipath iSCSI boot iPXE will wait until at least one path has become available and name resolution has completed on all other paths. We do this since the iBFT has to include IP addresses rather than DNS names. We will commence booting without waiting for the inactive paths to either become available or close; this avoids unnecessary boot delays. Note that the Linux kernel will refuse to accept an iBFT with more than two NIC or target structures. We therefore describe only the NICs that are actually required in order to reach the described targets. Any iBFT with at most two targets is therefore guaranteed to describe at most two NICs. Signed-off-by: Michael Brown <mcb30@ipxe.org>
2017-03-27 17:20:34 +02:00
/** iBFT model */
struct acpi_model ibft_model __acpi_model = {
.descs = LIST_HEAD_INIT ( ibft_model.descs ),
.complete = ibft_complete,
.install = ibft_install,
};