SeaBIOS

seabios@seabios.org

3 participants
3799 discussions

[PATCH] csm: Fix boot priority translation
by David Woodhouse 27 Jun '19

27 Jun '19

For CSM, the highest priority is zero. In SeaBIOS that means "don't", and the highest priority is 1. So we end up with the fun outcome that booting from NVMe worked only when it *wasn't* selected as the primary boot target, because we don't actually run the nvme_controller_setup() thread for an NVMe controller if its boot prio is zero. Signed-off-by: David Woodhouse <dwmw2(a)infradead.org> --- src/fw/csm.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) diff --git a/src/fw/csm.c b/src/fw/csm.c index 3fcc252..7663d31 100644 --- a/src/fw/csm.c +++ b/src/fw/csm.c @@ -319,6 +319,20 @@ handle_csm(struct bregs *regs) csm_return(regs); } +static int csm_prio_to_seabios(u16 csm_prio) +{ + switch (csm_prio) { + case BBS_DO_NOT_BOOT_FROM: + case BBS_IGNORE_ENTRY: + return -1; + + case BBS_LOWEST_PRIORITY: + case BBS_UNPRIORITIZED_ENTRY: + default: + return csm_prio + 1; + } +} + int csm_bootprio_ata(struct pci_device *pci, int chanid, int slave) { if (!csm_boot_table) @@ -327,7 +341,7 @@ int csm_bootprio_ata(struct pci_device *pci, int chanid, int slave) int index = 1 + (chanid * 2) + slave; dprintf(3, "CSM bootprio for ATA%d,%d (index %d) is %d\n", chanid, slave, index, bbs[index].BootPriority); - return bbs[index].BootPriority; + return csm_prio_to_seabios(bbs[index].BootPriority); } int csm_bootprio_fdc(struct pci_device *pci, int port, int fdid) @@ -336,7 +350,7 @@ int csm_bootprio_fdc(struct pci_device *pci, int port, int fdid) return -1; BBS_TABLE *bbs = (void *)csm_boot_table->BbsTable; dprintf(3, "CSM bootprio for FDC is %d\n", bbs[0].BootPriority); - return bbs[0].BootPriority; + return csm_prio_to_seabios(bbs[0].BootPriority); } int csm_bootprio_pci(struct pci_device *pci) @@ -350,7 +364,7 @@ int csm_bootprio_pci(struct pci_device *pci) if (pci->bdf == pci_to_bdf(bbs[i].Bus, bbs[i].Device, bbs[i].Function)) { dprintf(3, "CSM bootprio for PCI(%d,%d,%d) is %d\n", bbs[i].Bus, bbs[i].Device, bbs[i].Function, bbs[i].BootPriority); - return bbs[i].BootPriority; + return csm_prio_to_seabios(bbs[i].BootPriority); } } return -1;

4 11

[PATCH v3 0/4] Add Qemu to SeaBIOS LCHS interface
by Sam Eiderman 26 Jun '19

26 Jun '19

v1: Non-standard logical geometries break under QEMU. A virtual disk which contains an operating system which depends on logical geometries (consistent values being reported from BIOS INT13 AH=08) will most likely break under QEMU/SeaBIOS if it has non-standard logical geometries - for example 56 SPT (sectors per track). No matter what QEMU will guess - SeaBIOS, for large enough disks - will use LBA translation, which will report 63 SPT instead. In addition we can not enforce SeaBIOS to rely on phyiscal geometries at all. A virtio-blk-pci virtual disk with 255 phyiscal heads can not report more than 16 physical heads when moved to an IDE controller, the ATA spec allows a maximum of 16 heads - this is an artifact of virtualization. By supplying the logical geometies directly we are able to support such "exotic" disks. We will use fw_cfg to do just that. v2: Rename bootdevices fw_cfg key to bios-geoemtry v3: Change fw_cfg interface from mixed binary/textual to textual only Squash commit "config: Add toggle for bootdevice information" Sam Eiderman (4): geometry: Read LCHS from fw_cfg boot: Reorder functions in boot.c geometry: Add boot_lchs_find_*() utility functions geometry: Apply LCHS values for boot devices src/Kconfig | 7 ++ src/block.c | 21 ++++- src/block.h | 1 + src/boot.c | 239 +++++++++++++++++++++++++++++++++++++++++---------- src/hw/ahci.c | 1 + src/hw/ata.c | 8 ++ src/hw/esp-scsi.c | 2 + src/hw/lsi-scsi.c | 2 + src/hw/megasas.c | 1 + src/hw/mpt-scsi.c | 2 + src/hw/pvscsi.c | 1 + src/hw/virtio-blk.c | 2 + src/hw/virtio-scsi.c | 2 + src/util.h | 6 ++ 14 files changed, 249 insertions(+), 46 deletions(-) -- 2.13.3

3 14

Re: [QEMU] [PATCH v4 0/8] Add Qemu to SeaBIOS LCHS interface
by Gerd Hoffmann 20 Jun '19

20 Jun '19

> === OUTPUT BEGIN === > 1/8 Checking commit cf2168a66dbd (block: Refactor macros - fix tabbing) > ERROR: Macros with complex values should be enclosed in parenthesis > #55: FILE: include/hw/block/block.h:65: > +#define DEFINE_BLOCK_CHS_PROPERTIES(_state, _conf) \ > + DEFINE_PROP_UINT32("cyls", _state, _conf.cyls, 0), \ > + DEFINE_PROP_UINT32("heads", _state, _conf.heads, 0), \ > DEFINE_PROP_UINT32("secs", _state, _conf.secs, 0) False positive. > 8/8 Checking commit 6664d7d898d5 (hd-geo-test: Add tests for lchs override) > WARNING: line over 80 characters > #242: FILE: tests/hd-geo-test.c:579: > + (char *)r->dev_path, &(r->chs.c), &(r->chs.h), &(r->chs.s)); Should be fixed. > WARNING: Block comments use a leading /* on a separate line > #640: FILE: tests/hd-geo-test.c:995: > + "skipping hd-geo/override/* tests"); False positive too. cheers, Gerd

2 1

[QEMU] [PATCH v4 0/8] Add Qemu to SeaBIOS LCHS interface
by Sam Eiderman 20 Jun '19

20 Jun '19

v1: Non-standard logical geometries break under QEMU. A virtual disk which contains an operating system which depends on logical geometries (consistent values being reported from BIOS INT13 AH=08) will most likely break under QEMU/SeaBIOS if it has non-standard logical geometries - for example 56 SPT (sectors per track). No matter what QEMU will guess - SeaBIOS, for large enough disks - will use LBA translation, which will report 63 SPT instead. In addition we can not enforce SeaBIOS to rely on phyiscal geometries at all. A virtio-blk-pci virtual disk with 255 phyiscal heads can not report more than 16 physical heads when moved to an IDE controller, the ATA spec allows a maximum of 16 heads - this is an artifact of virtualization. By supplying the logical geometies directly we are able to support such "exotic" disks. We will use fw_cfg to do just that. v2: Fix missing parenthesis check in "hd-geo-test: Add tests for lchs override" v3: * Rename fw_cfg key to "bios-geometry". * Remove "extendible" interface. * Add cpu_to_le32 fix as Laszlo suggested or big endian hosts * Fix last qtest commit - automatic docker tester for some reason does not have qemu-img set v4: * Change fw_cfg interface from mixed textual/binary to textual only Sam Eiderman (8): block: Refactor macros - fix tabbing block: Support providing LCHS from user bootdevice: Add interface to gather LCHS scsi: Propagate unrealize() callback to scsi-hd bootdevice: Gather LCHS from all relevant devices bootdevice: Refactor get_boot_devices_list bootdevice: FW_CFG interface for LCHS values hd-geo-test: Add tests for lchs override bootdevice.c | 148 +++++++++--- hw/block/virtio-blk.c | 6 + hw/ide/qdev.c | 7 +- hw/nvram/fw_cfg.c | 14 +- hw/scsi/scsi-bus.c | 15 ++ hw/scsi/scsi-disk.c | 14 ++ include/hw/block/block.h | 22 +- include/hw/scsi/scsi.h | 1 + include/sysemu/sysemu.h | 4 + tests/Makefile.include | 2 +- tests/hd-geo-test.c | 581 +++++++++++++++++++++++++++++++++++++++++++++++ 11 files changed, 773 insertions(+), 41 deletions(-) -- 2.13.3

2 9

[PATCH] scripts/buildversion.py: allow version to be overridden
by David Woodhouse 19 Jun '19

19 Jun '19

Signed-off-by: David Woodhouse <dwmw2(a)infradead.org> --- scripts/buildversion.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/scripts/buildversion.py b/scripts/buildversion.py index 8875497..fc2decd 100755 --- a/scripts/buildversion.py +++ b/scripts/buildversion.py @@ -113,7 +113,9 @@ def main(): cleanbuild, toolstr = tool_versions(options.tools) - ver = git_version() + ver = os.getenv('SEABIOS_VERSION') + if not ver: + ver = git_version() cleanbuild = cleanbuild and 'dirty' not in ver if not ver: ver = file_version()

2 4

[PATCH] csm: Sanitise alignment constraint in Legacy16GetTableAddress
by David Woodhouse 18 Jun '19

18 Jun '19

The alignment constraint is defined in the CSM specifications as "Bit mapped. First non-zero bit from the right is the alignment." Use __fls() to sanitise the alignment given that definition, since passing a non-power-of-two alignment to _malloc() isn't going to work well. And cope with being passed zero, which was happening for the E820 table allocation from EDK2. Signed-off-by: David Woodhouse <dwmw2(a)infradead.org> --- src/fw/csm.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/src/fw/csm.c b/src/fw/csm.c index 03b4bb8..bf7b8f5 100644 --- a/src/fw/csm.c +++ b/src/fw/csm.c @@ -264,6 +264,13 @@ handle_csm_0006(struct bregs *regs) dprintf(3, "Legacy16GetTableAddress size %x align %x region %d\n", size, align, region); + // DX = Required address alignment. Bit mapped. + // First non-zero bit from the right is the alignment.*/ + if (align) + align = 1 << __ffs(align); + else + align = 1; + if (region & 2) chunk = _malloc(&ZoneLow, size, align); if (!chunk && (region & 1))

2 3

[QEMU] [PATCH v2 0/8] Add Qemu to SeaBIOS LCHS interface
by Sam Eiderman 17 Jun '19

17 Jun '19

v1: Non-standard logical geometries break under QEMU. A virtual disk which contains an operating system which depends on logical geometries (consistent values being reported from BIOS INT13 AH=08) will most likely break under QEMU/SeaBIOS if it has non-standard logical geometries - for example 56 SPT (sectors per track). No matter what QEMU will guess - SeaBIOS, for large enough disks - will use LBA translation, which will report 63 SPT instead. In addition we can not enforce SeaBIOS to rely on phyiscal geometries at all. A virtio-blk-pci virtual disk with 255 phyiscal heads can not report more than 16 physical heads when moved to an IDE controller, the ATA spec allows a maximum of 16 heads - this is an artifact of virtualization. By supplying the logical geometies directly we are able to support such "exotic" disks. We will use fw_cfg to do just that. v2: Fix missing parenthesis check in "hd-geo-test: Add tests for lchs override" Sam Eiderman (8): block: Refactor macros - fix tabbing block: Support providing LCHS from user bootdevice: Add interface to gather LCHS scsi: Propagate unrealize() callback to scsi-hd bootdevice: Gather LCHS from all relevant devices bootdevice: Refactor get_boot_devices_list bootdevice: FW_CFG interface for LCHS values hd-geo-test: Add tests for lchs override bootdevice.c | 158 ++++++++++--- hw/block/virtio-blk.c | 6 + hw/ide/qdev.c | 7 +- hw/nvram/fw_cfg.c | 14 +- hw/scsi/scsi-bus.c | 15 ++ hw/scsi/scsi-disk.c | 14 ++ include/hw/block/block.h | 22 +- include/hw/scsi/scsi.h | 1 + include/sysemu/sysemu.h | 4 + tests/Makefile.include | 2 +- tests/hd-geo-test.c | 565 +++++++++++++++++++++++++++++++++++++++++++++++ 11 files changed, 767 insertions(+), 41 deletions(-) -- 2.13.3

3 25

[QEMU] [PATCH v3 0/8] Add Qemu to SeaBIOS LCHS interface
by Sam Eiderman 17 Jun '19

17 Jun '19

v1: Non-standard logical geometries break under QEMU. A virtual disk which contains an operating system which depends on logical geometries (consistent values being reported from BIOS INT13 AH=08) will most likely break under QEMU/SeaBIOS if it has non-standard logical geometries - for example 56 SPT (sectors per track). No matter what QEMU will guess - SeaBIOS, for large enough disks - will use LBA translation, which will report 63 SPT instead. In addition we can not enforce SeaBIOS to rely on phyiscal geometries at all. A virtio-blk-pci virtual disk with 255 phyiscal heads can not report more than 16 physical heads when moved to an IDE controller, the ATA spec allows a maximum of 16 heads - this is an artifact of virtualization. By supplying the logical geometies directly we are able to support such "exotic" disks. We will use fw_cfg to do just that. v2: Fix missing parenthesis check in "hd-geo-test: Add tests for lchs override" v3: * Rename fw_cfg key to "bios-geometry". * Remove "extendible" interface. * Add cpu_to_le32 fix as Laszlo suggested or big endian hosts * Fix last qtest commit - automatic docker tester for some reason does not have qemu-img set Sam Eiderman (8): block: Refactor macros - fix tabbing block: Support providing LCHS from user bootdevice: Add interface to gather LCHS scsi: Propagate unrealize() callback to scsi-hd bootdevice: Gather LCHS from all relevant devices bootdevice: Refactor get_boot_devices_list bootdevice: FW_CFG interface for LCHS values hd-geo-test: Add tests for lchs override bootdevice.c | 154 ++++++++++--- hw/block/virtio-blk.c | 6 + hw/ide/qdev.c | 7 +- hw/nvram/fw_cfg.c | 14 +- hw/scsi/scsi-bus.c | 15 ++ hw/scsi/scsi-disk.c | 14 ++ include/hw/block/block.h | 22 +- include/hw/scsi/scsi.h | 1 + include/sysemu/sysemu.h | 4 + tests/Makefile.include | 2 +- tests/hd-geo-test.c | 566 +++++++++++++++++++++++++++++++++++++++++++++++ 11 files changed, 764 insertions(+), 41 deletions(-) -- 2.13.3

1 8

[PATCH v2 0/5] Add Qemu to SeaBIOS LCHS interface
by Sam Eiderman 17 Jun '19

17 Jun '19

v1: Non-standard logical geometries break under QEMU. A virtual disk which contains an operating system which depends on logical geometries (consistent values being reported from BIOS INT13 AH=08) will most likely break under QEMU/SeaBIOS if it has non-standard logical geometries - for example 56 SPT (sectors per track). No matter what QEMU will guess - SeaBIOS, for large enough disks - will use LBA translation, which will report 63 SPT instead. In addition we can not enforce SeaBIOS to rely on phyiscal geometries at all. A virtio-blk-pci virtual disk with 255 phyiscal heads can not report more than 16 physical heads when moved to an IDE controller, the ATA spec allows a maximum of 16 heads - this is an artifact of virtualization. By supplying the logical geometies directly we are able to support such "exotic" disks. We will use fw_cfg to do just that. v2: Rename bootdevices fw_cfg key to bios-geoemtry Sam Eiderman (5): geometry: Read LCHS from fw_cfg boot: Reorder functions in boot.c geometry: Add boot_lchs_find_*() utility functions config: Add toggle for bootdevice information geometry: Apply LCHS values for boot devices src/Kconfig | 7 ++ src/block.c | 21 ++++- src/block.h | 1 + src/boot.c | 244 +++++++++++++++++++++++++++++++++++++++++---------- src/hw/ahci.c | 1 + src/hw/ata.c | 8 ++ src/hw/esp-scsi.c | 2 + src/hw/lsi-scsi.c | 2 + src/hw/megasas.c | 1 + src/hw/mpt-scsi.c | 2 + src/hw/pvscsi.c | 1 + src/hw/virtio-blk.c | 2 + src/hw/virtio-scsi.c | 2 + src/util.h | 6 ++ 14 files changed, 254 insertions(+), 46 deletions(-) -- 2.13.3

1 5

BUG: Seabios choose the wrong option rom when two are available
by akjuxr3＠dismail.de 17 Jun '19

17 Jun '19

In this thread https://mail.coreboot.org/hyperkitty/list/coreboot@coreboot.org/thread/FTHN… we found a bug in seabios. Its choosing the wrong option rom in the case when for example on a Intel G41 board you have the internal Intel GPU and have added a external PCIe GPU. Then coreboot switches over to the external GPU for the graphical output but seabios use the option rom for the Intel G41 GPU. In case of x4x/G41 this results in no ability to have a graphical boot mode and thus its not possible to boot some linux live images. In the linked thread a workaround is been found: "In the "Devices" menu of coreboot, set "Graphics initialization" to "None" ". This breaks the functionality to have coreboot output with the internal GPU when you pull out the PCIe GPU. The switch should happen automaticly and without the need to recompile/reconfigure coreboot. Changes like https://review.coreboot.org/c/coreboot/+/18504 was made to make this happen. Would be great if this seabios issue could be fixes. Thanks!

2 1

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

SeaBIOS