SeaBIOS October 2011

seabios@seabios.org

33 participants
56 discussions

[PATCH 0/5] More pci init simplifications
by Kevin O'Connor 30 Oct '11

30 Oct '11

This main patch in this series is patch 2 - replacement of the recursive scan with a linear scan. The other patches are just cosmetic changes. Gerd - can you take a look at patch 2. Also, what's your test case for confirming multiple bus initialization? The standard qemu/kvm only has one host bus. -Kevin Kevin O'Connor (5): Replace pciinit busses_count with MaxPCIBus. Replace recursive pci init system with linear passes. Locally allocate pciinit busses[] variable. Move pciinit device init code together. Simplify pci_bios_init_root_regions(). src/pciinit.c | 245 +++++++++++++++++++++++++-------------------------------- 1 files changed, 107 insertions(+), 138 deletions(-) -- 1.7.6.4

3 10

[PATCH 1/2] seabios: Add Local APIC NMI Structure to ACPI MADT (was: Re: [Qemu-devel] [PATCH] qemu: Fix inject-nmi)
by Lai Jiangshan 30 Oct '11

30 Oct '11

From: Kenji Kaneshige <kaneshige.kenji(a)jp.fujitsu.com> ACPI NMI Structure describes LINT pin (LINT0 or LINT1) information to which NMI is connected, and it is needed by OS to initialize local APIC. Signed-off-by: Kenji Kaneshige <kaneshige.kenji(a)jp.fujitsu.com> Reviewed-by: Lai Jiangshan <laijs(a)cn.fujitsu.com> --- src/acpi.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-) Index: seabios/src/acpi.c =================================================================== --- seabios.orig/src/acpi.c +++ seabios/src/acpi.c @@ -134,6 +134,14 @@ struct madt_intsrcovr { u16 flags; } PACKED; +struct madt_local_nmi { + ACPI_SUB_HEADER_DEF + u8 processor_id; /* ACPI processor id */ + u16 flags; /* MPS INTI flags */ + u8 lint; /* Local APIC LINT# */ +} PACKED; + + /* * ACPI 2.0 Generic Address Space definition. */ @@ -288,7 +296,9 @@ build_madt(void) int madt_size = (sizeof(struct multiple_apic_table) + sizeof(struct madt_processor_apic) * MaxCountCPUs + sizeof(struct madt_io_apic) - + sizeof(struct madt_intsrcovr) * 16); + + sizeof(struct madt_intsrcovr) * 16 + + sizeof(struct madt_local_nmi)); + struct multiple_apic_table *madt = malloc_high(madt_size); if (!madt) { warn_noalloc(); @@ -340,7 +350,15 @@ build_madt(void) intsrcovr++; } - build_header((void*)madt, APIC_SIGNATURE, (void*)intsrcovr - (void*)madt, 1); + struct madt_local_nmi *local_nmi = (void*)intsrcovr; + local_nmi->type = APIC_LOCAL_NMI; + local_nmi->length = sizeof(*local_nmi); + local_nmi->processor_id = 0xff; /* all processors */ + local_nmi->flags = 0; + local_nmi->lint = 1; /* LINT1 */ + local_nmi++; + + build_header((void*)madt, APIC_SIGNATURE, (void*)local_nmi - (void*)madt, 1); return madt; }

5 5

Re: [SeaBIOS] [PATCH] Set numa topology for max_cpus
by Kevin O'Connor 28 Oct '11

28 Oct '11

On Wed, Oct 26, 2011 at 02:19:00PM +0200, Vasilis Liaskovitis wrote: > qemu-kvm passes numa/SRAT topology information for smp_cpus to SeaBIOS. However > SeaBIOS always expects to setup max_cpus number of SRAT cpu entries > (MaxCountCPUs variable in build_srat function of Seabios). When qemu-kvm runs > with smp_cpus != max_cpus (e.g. -smp 2,maxcpus=4), Seabios will mistakenly use > memory SRAT info for setting up CPU SRAT entries for the offline CPUs. Wrong > SRAT memory entries are also created. This breaks NUMA in a guest. > Fix by setting up SRAT info for max_cpus in qemu-kvm. > > Signed-off-by: Vasilis Liaskovitis <vasilis.liaskovitis(a)profitbricks.com> It looks okay to me, but I'd like to see an Ack from a QEmu or KVM maintainer. -Kevin

1 0

Re: [SeaBIOS] [Xen-devel] Question on virtual block device
by Ian Campbell 27 Oct '11

27 Oct '11

On Thu, 2011-10-27 at 17:14 +0100, Daniel Castro wrote: > Hello All, > > I am connecting my front and back for a block device. > I have collected from xenstore sectors, sector-size. But I also need: > u16 heads; // # heads > u16 cylinders; // # cylinders > u16 spt; // # sectors / track > I guess from the SeaBIOS perspective they are required, but from the > xen perspective how can I get them? pure math/translation? CHS is rather historical and has no real relevance to a virtual device. You can basically just make them up to fit the underlying device size, take a look at blkif_getgeo in Linux for example. Ian.

1 0

Question on virtual block device
by Daniel Castro 27 Oct '11

27 Oct '11

Hello All, I am connecting my front and back for a block device. I have collected from xenstore sectors, sector-size. But I also need: u16 heads; // # heads u16 cylinders; // # cylinders u16 spt; // # sectors / track I guess from the SeaBIOS perspective they are required, but from the xen perspective how can I get them? pure math/translation? Thanks, Daniel -- +-=====---------------------------+ | +---------------------------------+ | This space intentionally blank for notetaking. | | | Daniel Castro, | | | | Consultant/Programmer.| | | | U Andes | +-------------------------------------+

1 0

Re: [SeaBIOS] Status of SeaBIOS VGA bios.
by Kevin O'Connor 27 Oct '11

27 Oct '11

On Wed, Oct 26, 2011 at 02:06:53AM +0100, Julian Pidancet wrote: > On Wed, Oct 26, 2011 at 1:26 AM, Kevin O'Connor <kevin(a)koconnor.net> wrote: > > It should work. Ideally, someone would complete the port of the "lgpl > > vgabios". This involves translating the remaining 16bit assembler to > > C code and then doing lots of testing on the result. > > I havn't tried to run the ROM yet. Could you be more precise about > what works (or what is supposed to be working), what's not, and what > exactly remains to be translated ? I'd like to make an idea of the > amount of work required to have a ROM equivalent to vgabios. If you look at the "lgpl vgabios" source - I ported vgabios.c but did not convert vbe.c nor clext.c. As for what works - I was able to launch qemu and run guests. However, the vbe extensions weren't present, and so most guests wouldn't attempt to increase resolution or they had other features not available. Also, I was seeing the grub screen come up as black-and-white - which may be due to lack of VBE, but may also have been due to a bug somewhere in the port. -Kevin

2 1

Re: [SeaBIOS] git http access to repository
by Peter Stuge 27 Oct '11

27 Oct '11

Ram Bhamidipaty wrote: > Hi - I am behind a corporate firewall and cannot directly access git > repositories. Is is possible to access the seabios git repository > via the http protocol? Why not just try it? $ git clone http://git.seabios.org/seabios.git Cloning into seabios... fatal: http://git.seabios.org/seabios.git/info/refs not found: did you run git update-server-info on the server? $ You could have sent the error message already. I guess this will be fixed fairly quickly. //Peter

1 0

git http access to repository
by Ram Bhamidipaty 26 Oct '11

26 Oct '11

Hi - I am behind a corporate firewall and cannot directly access git repositories. Is is possible to access the seabios git repository via the http protocol? Thank you for any info. -Ram

1 0

[PATCH] Set numa topology for max_cpus
by Vasilis Liaskovitis 26 Oct '11

26 Oct '11

qemu-kvm passes numa/SRAT topology information for smp_cpus to SeaBIOS. However SeaBIOS always expects to setup max_cpus number of SRAT cpu entries (MaxCountCPUs variable in build_srat function of Seabios). When qemu-kvm runs with smp_cpus != max_cpus (e.g. -smp 2,maxcpus=4), Seabios will mistakenly use memory SRAT info for setting up CPU SRAT entries for the offline CPUs. Wrong SRAT memory entries are also created. This breaks NUMA in a guest. Fix by setting up SRAT info for max_cpus in qemu-kvm. Signed-off-by: Vasilis Liaskovitis <vasilis.liaskovitis(a)profitbricks.com> --- hw/pc.c | 8 ++++---- vl.c | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/hw/pc.c b/hw/pc.c index fb49612..e0a65da 100644 --- a/hw/pc.c +++ b/hw/pc.c @@ -626,9 +626,9 @@ static void *bochs_bios_init(void) * of nodes, one word for each VCPU->node and one word for each node to * hold the amount of memory. */ - numa_fw_cfg = g_malloc0((1 + smp_cpus + nb_numa_nodes) * 8); + numa_fw_cfg = g_malloc0((1 + max_cpus + nb_numa_nodes) * 8); numa_fw_cfg[0] = cpu_to_le64(nb_numa_nodes); - for (i = 0; i < smp_cpus; i++) { + for (i = 0; i < max_cpus; i++) { for (j = 0; j < nb_numa_nodes; j++) { if (node_cpumask[j] & (1 << i)) { numa_fw_cfg[i + 1] = cpu_to_le64(j); @@ -637,10 +637,10 @@ static void *bochs_bios_init(void) } } for (i = 0; i < nb_numa_nodes; i++) { - numa_fw_cfg[smp_cpus + 1 + i] = cpu_to_le64(node_mem[i]); + numa_fw_cfg[max_cpus + 1 + i] = cpu_to_le64(node_mem[i]); } fw_cfg_add_bytes(fw_cfg, FW_CFG_NUMA, (uint8_t *)numa_fw_cfg, - (1 + smp_cpus + nb_numa_nodes) * 8); + (1 + max_cpus + nb_numa_nodes) * 8); return fw_cfg; } diff --git a/vl.c b/vl.c index aeed54a..a3ae262 100644 --- a/vl.c +++ b/vl.c @@ -3410,7 +3410,7 @@ int main(int argc, char **argv, char **envp) * real machines which also use this scheme. */ if (i == nb_numa_nodes) { - for (i = 0; i < smp_cpus; i++) { + for (i = 0; i < max_cpus; i++) { node_cpumask[i % nb_numa_nodes] |= 1 << i; } } -- 1.7.6.3

1 0

Re: [SeaBIOS] Status of SeaBIOS VGA bios.
by Kevin O'Connor 26 Oct '11

26 Oct '11

On Wed, Oct 26, 2011 at 12:16:32AM +0100, Julian Pidancet wrote: > Hi, > > I stumbled across the vgasrc directory in the SeaBIOS git tree which > seem to contain the source of a VGA Bios. I havn't found any related > documentation nor announcements regarding this, and there is no > Makefiles or scripts to build it. That's why I'm enquiring here about > the status of this project. You can build the vgabios by running: make out/vgabios.bin There was a brief note on the vga code some time back: http://www.coreboot.org/pipermail/coreboot/2009-May/047775.html > I've got several questions: > - What is the general status of this project ? I started converting the VGA bios from the "lgpl vga bios" project ( http://savannah.nongnu.org/projects/vgabios/ ) a couple of years ago. I converted the base code, but did not have time to convert the "VBE" or the "cirrus" code. There was also an effort a few years back to get the code running on real hardware that had some success. > - Is this code intended to be executed in a 32bit environment by > transitionning from 16bit to 32bit mode for each BIOS call as SeaBIOS > already does ? The current seabios vgabios code runs exclusively in 16bit mode. (Note, SeaBIOS also has some calls handled entirely in 16bit mode as well.) > - Is this code able to run independantly from SeaBIOS ? It should - but I've never tested it that way. > - How can I build the sources ? make out/vgabios.bin > I am currently working on a emulated graphic adapter for Xen and I > would like to implement a VBE driver for it. This VGA BIOS seems to be > a good base for that kind of work. Would that be possible ? It should work. Ideally, someone would complete the port of the "lgpl vgabios". This involves translating the remaining 16bit assembler to C code and then doing lots of testing on the result. -Kevin

3 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

SeaBIOS October 2011