SeaBIOS

seabios@seabios.org

3783 discussions

Re: [SeaBIOS] A20: ahci_process_op is called with A20 off
by Kevin O'Connor 13 May '17

13 May '17

On Thu, May 11, 2017 at 11:22:02PM +0000, Xu, Anthony wrote: > > SeaBIOS has a couple of different methods to accomplish this mode > > switching - it can directly switch modes (C16_BIG switch) or it can > > use a helper in SMM mode to perform the switch (C16_SMM). The > > preferred method is C16_SMM as C16_BIG isn't available in VM86 mode > > and it also clobbers the 16bit segment registers. > > Hi Kevin, > > Why does SeaBios use VM86 mode? As Paolo mentioned, SeaBIOS doesn't use it, but it must support callers that use it. > vtx supports real mode long time ago. > > C16_SMM is slow because it causes several extra switches between VM and QEMU. > I would prefer C16_BIG for better performance. > Does SeaBios know if it runs in real mode or VM86 mode? > Can SeaBios choose C16_SMM/C16_BIG based on VCPU mode? Yes, SeaBIOS could use the smsw instruction (see x86.h:cr0_vm86_read) to check if it is in VM86 mode. Using C16_BIG would still clobber the segment registers, but it might be worth that risk to avoid the extra overhead of C16_SMM. -Kevin

3 3

Re: [SeaBIOS] A20: ahci_process_op is called with A20 off
by Paolo Bonzini 12 May '17

12 May '17

On 12/05/2017 01:22, Xu, Anthony wrote: >> SeaBIOS has a couple of different methods to accomplish this mode >> switching - it can directly switch modes (C16_BIG switch) or it can >> use a helper in SMM mode to perform the switch (C16_SMM). The >> preferred method is C16_SMM as C16_BIG isn't available in VM86 mode >> and it also clobbers the 16bit segment registers. > > Hi Kevin, > > Why does SeaBios use VM86 mode? SeaBIOS doesn't, but for example old DOS extenders uses it. You can run Windows 95 inside SeaBIOS. Paolo > vtx supports real mode long time ago. > > C16_SMM is slow because it causes several extra switches between VM and QEMU. > I would prefer C16_BIG for better performance. > Does SeaBios know if it runs in real mode or VM86 mode? > Can SeaBios choose C16_SMM/C16_BIG based on VCPU mode? > > > Thanks, > Anthony >

1 0

Re: [SeaBIOS] A20: ahci_process_op is called with A20 off
by Kevin O'Connor 12 May '17

12 May '17

On Wed, May 10, 2017 at 07:28:44PM +0000, Xu, Anthony wrote: > > From: Kevin O'Connor [mailto:kevin@koconnor.net] > > On Tue, May 09, 2017 at 08:39:07PM +0000, Xu, Anthony wrote: > > > I dig into seebios code, and found ahci_process_op is called with A20 off, > > > > > > The call trace is, > > > process_op_16->process_op_both->call32->call32_smm->process_op_32- > > >ahci_process_op > > > > > > call32_smm calls call32_prep, but because the method is C16_SMM, A20 is > > not enabled in call32_prep. > > > ahci_process_op is called with A20 off. > > > > As I understand it, entering System Managament Mode (SMM) should > > automatically enable A20, so it should not be necessary for call32_smm > > to enable it. Are you sure the issue is related to A20 instead of > > some other issue with SMM? > > I'm pretty sure the issue is related to A20, I'm not familiar with Seabios code, > I'm not sure why call32 needs to call32_smm. If my understanding is correct, > VCPU is not in SMM mode at the time. As background, seabios must support disk read/write requests when the x86 processor is in 16bit mode. There's some info on the modes seabios supports at: https://www.seabios.org/Memory_Model Unfortunately, some disk controllers can only be controlled when the processor is in 32bit mode - ahci is an example of this. So, to service these calls, seabios must switch from 16bit mode to 32bit mode, perform the disk operation, and then return to 16bit mode. SeaBIOS has a couple of different methods to accomplish this mode switching - it can directly switch modes (C16_BIG switch) or it can use a helper in SMM mode to perform the switch (C16_SMM). The preferred method is C16_SMM as C16_BIG isn't available in VM86 mode and it also clobbers the 16bit segment registers. If you don't want to use C16_SMM, you can turn off CONFIG_CALL32_SMM via "make menuconfig". It's still not clear to me why C16_SMM doesn't work though - as I think it should have saved and restored the A20 flag. -Kevin

2 1

Re: [SeaBIOS] A20: ahci_process_op is called with A20 off
by Kevin O'Connor 11 May '17

11 May '17

On Tue, May 09, 2017 at 08:39:07PM +0000, Xu, Anthony wrote: > Hi All, > > I recently met an issue, I can't boot my centos guest in below scenarios. > > I disabled kvmvapic by > > diff --git a/hw/intc/apic_common.c b/hw/intc/apic_common.c > index c3829e3..52be2b0 100644 > --- a/hw/intc/apic_common.c > +++ b/hw/intc/apic_common.c > @@ -440,7 +440,7 @@ static const VMStateDescription vmstate_apic_common = { > static Property apic_properties_common[] = { > DEFINE_PROP_UINT8("version", APICCommonState, version, 0x14), > DEFINE_PROP_BIT("vapic", APICCommonState, vapic_control, VAPIC_ENABLE_BIT, > - true), > + false), > DEFINE_PROP_BOOL("legacy-instance-id", APICCommonState, legacy_instance_id, > false), > DEFINE_PROP_END_OF_LIST(), > > > I run Qemu in TCG mode > > ./x86_64-softmmu/qemu-system-x86_64 -bios /home/root/git/seabios/out/bios.bin -machine q35 > -m 1G -drive format=raw,file=/home/root/images/centos7.2.img,if=ide,index=0 -nographic > -nodefaults -serial stdio -monitor pty > > It can't boot with below message, > " > Booting from Hard Disk... > WARNING - Timeout at ahci_command:154! > Boot failed: could not read the boot disk > > enter handle_18: > NULL > Booting from Floppy... > Boot failed: could not read the boot disk > " > > Once I add kvmvapic , it can boot. > > I dig into seebios code, and found ahci_process_op is called with A20 off, > > The call trace is, > process_op_16->process_op_both->call32->call32_smm->process_op_32->ahci_process_op > > call32_smm calls call32_prep, but because the method is C16_SMM, A20 is not enabled in call32_prep. > ahci_process_op is called with A20 off. As I understand it, entering System Managament Mode (SMM) should automatically enable A20, so it should not be necessary for call32_smm to enable it. Are you sure the issue is related to A20 instead of some other issue with SMM? -Kevin

2 1

[PATCH] NVMe: Allow NVMe to be enabled on real hardware
by Kevin O'Connor 11 May '17

11 May '17

A couple of users have reported success with the NVMe driver on real hardware, so allow it to be enabled outside of QEMU. Signed-off-by: Kevin O'Connor <kevin(a)koconnor.net> --- src/Kconfig | 2 +- src/hw/nvme.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/Kconfig b/src/Kconfig index e1b83a4..77ec9c7 100644 --- a/src/Kconfig +++ b/src/Kconfig @@ -228,7 +228,7 @@ menu "Hardware support" Support floppy images stored in coreboot flash or from QEMU fw_cfg. config NVME - depends on DRIVES && QEMU_HARDWARE + depends on DRIVES bool "NVMe controllers" default y help diff --git a/src/hw/nvme.c b/src/hw/nvme.c index 60557ae..1c2bce5 100644 --- a/src/hw/nvme.c +++ b/src/hw/nvme.c @@ -618,7 +618,7 @@ void nvme_setup(void) { ASSERT32FLAT(); - if (!CONFIG_NVME || !runningOnQEMU()) + if (!CONFIG_NVME) return; dprintf(3, "init nvme\n"); -- 2.9.3

1 0

Re: [SeaBIOS] [PATCH 1/4] SeaVGABios/cbvga: Assume VGA compatible GPU in text-mode
by Patrick Rudolph 11 May '17

11 May '17

On Sat Mar 18 02:04:38 CET 2017, Kevin O'Connor wrote: >> Assume the GPU is VGA compatible when running in text-mode. >> Advertise VGA modes using SeaVGABios' stdvga driver. >> running in text mode, even when not advertised. >> >> Allows Windows 7 to boot in 640x480 and 16 colors mode using the >> VgaSave driver. >> Allows Windows 7 to show the loading bar splash screen. >> >> Known issues: >> The palette seems wrong, as colors are swapped ( blue is red, ... ). > >What graphics hardware did you test this on? > >If stdvga_set_mode() works on the hardware, why not use stdvga >seavgabios and get the rest of the vgabios functionality? (That is, >why use cbvga seavgabios at all?) I'm on Lenovo T500 (Intel Graphics Media Accelerator 4500MHD). I'm going to do some tests using stdvga instead of cbvga. Regards, Patrick

2 1

Re: [SeaBIOS] [Question] SeaBIOS: cannot boot oracle linux if increase the maximum size of permanent high memory area
by Kevin O'Connor 11 May '17

11 May '17

On Fri, May 05, 2017 at 10:49:05AM +0000, Gonglei (Arei) wrote: > Hi guys, > > Currently my workmate encountered an issues in the testing environment: > > A letter from him: > > In order to boot a BIG vm (with 4T mem, 255 vCPUs, 60 virtio-scsi > disk...), i have to increase the BUILD_MAX_HIGHTABLE to 512KB. But, > then i found i can not boot a specific VM anymore (oracle linux 6.7 > 64bits with kernel 3.8.13, 2G mem). It seems be related with the > unused high ram given VM back by SeaBios (maybe also related to the > memory align). I've tested if setting giveback memory 64KB > alignment other than 4KB, SeaBios with 512KB BUILD_MAX_HIGHTABLE can > boot the specific VM. But , i don't know the reason... > > As far as i know, if seabios does not give back the unused ZoneHigh > memory seems will not trigger any problems but just waste only a > little memory ,right? > > Can you tell me what's the worst effect if i apply the following patch? > > Does it influence live migration? The only downside I'm aware of is that it would waste a small amount of ram. I'm surprised that "oracle linux" has this issue. I can only see a few reasons for it to fail there - a "use after free" bug in seabios, oracle linux expects the ram to be zero'd, or oracle linux is expecting the e820 map to have an alignment larger than page size. The first seems unlikely as "permanent high ram" is rarely ever free'd in seabios. The second also seems unlikely. -Kevin

1 0

A20: ahci_process_op is called with A20 off
by Xu, Anthony 10 May '17

10 May '17

Hi All, I recently met an issue, I can't boot my centos guest in below scenarios. I disabled kvmvapic by diff --git a/hw/intc/apic_common.c b/hw/intc/apic_common.c index c3829e3..52be2b0 100644 --- a/hw/intc/apic_common.c +++ b/hw/intc/apic_common.c @@ -440,7 +440,7 @@ static const VMStateDescription vmstate_apic_common = { static Property apic_properties_common[] = { DEFINE_PROP_UINT8("version", APICCommonState, version, 0x14), DEFINE_PROP_BIT("vapic", APICCommonState, vapic_control, VAPIC_ENABLE_BIT, - true), + false), DEFINE_PROP_BOOL("legacy-instance-id", APICCommonState, legacy_instance_id, false), DEFINE_PROP_END_OF_LIST(), I run Qemu in TCG mode ./x86_64-softmmu/qemu-system-x86_64 -bios /home/root/git/seabios/out/bios.bin -machine q35 -m 1G -drive format=raw,file=/home/root/images/centos7.2.img,if=ide,index=0 -nographic -nodefaults -serial stdio -monitor pty It can't boot with below message, " Booting from Hard Disk... WARNING - Timeout at ahci_command:154! Boot failed: could not read the boot disk enter handle_18: NULL Booting from Floppy... Boot failed: could not read the boot disk " Once I add kvmvapic , it can boot. I dig into seebios code, and found ahci_process_op is called with A20 off, The call trace is, process_op_16->process_op_both->call32->call32_smm->process_op_32->ahci_process_op call32_smm calls call32_prep, but because the method is C16_SMM, A20 is not enabled in call32_prep. ahci_process_op is called with A20 off. ahci_process_op accesses memory mapped IO to operate ahci, the IO is mapped to 0xfebff000, because a20 is disabled, the bit 20 of 0xfebff000 is set 0 , which is 0xfeaff000, an iotlb entry mapping 0xfebff000 to 0xfeaff000 is added to iotlb cache. So ahci_process_op actually accesses 0xfeaff000, it obviously won't work. target/i386/helper.c do_mapping: pte = pte & env->a20_mask; Why does guest boot with kvmvapic? In above scenario, there is no option roms. After adding kvmvapic, there is one option rom kvmvapic rom, and __callrom is called for kvmvapic rom. __callrom->farcall16big-> call16_override, once call16_override is called, a20 is enabled and will not be disabled. So ahci_process_op is called with a20 on. I tried below change, guest can boot without kvmvapic. I'm not familiar with seabios code, I don't consider it as a patch, but it proves my analysis is correct. With the change, __call32 calls call32_prep with method C16_BIG to enable a20 before calling ahci_process_op. please help fix this issue. Thanks, Anthony diff --git a/src/stacks.c b/src/stacks.c index ef6a707..0949652 100644 --- a/src/stacks.c +++ b/src/stacks.c @@ -232,10 +237,11 @@ u32 VISIBLE16 __call32(void *func, u32 eax, u32 errret) { ASSERT16(); - if (CONFIG_CALL32_SMM && GET_GLOBAL(HaveSmmCall32)) - return call32_smm(func, eax); +// if (CONFIG_CALL32_SMM && GET_GLOBAL(HaveSmmCall32)) +// return call32_smm(func, eax); // Jump direclty to 32bit mode - this clobbers the 16bit segment // selector registers.

1 0

[Question] SeaBIOS: cannot boot oracle linux if increase the maximum size of permanent high memory area
by Gonglei (Arei) 05 May '17

05 May '17

Hi guys, Currently my workmate encountered an issues in the testing environment: A letter from him: In order to boot a BIG vm (with 4T mem, 255 vCPUs, 60 virtio-scsi disk...), i have to increase the BUILD_MAX_HIGHTABLE to 512KB. But, then i found i can not boot a specific VM anymore (oracle linux 6.7 64bits with kernel 3.8.13, 2G mem). It seems be related with the unused high ram given VM back by SeaBios (maybe also related to the memory align). I've tested if setting giveback memory 64KB alignment other than 4KB, SeaBios with 512KB BUILD_MAX_HIGHTABLE can boot the specific VM. But , i don't know the reason... As far as i know, if seabios does not give back the unused ZoneHigh memory seems will not trigger any problems but just waste only a little memory ,right? Can you tell me what's the worst effect if i apply the following patch? Does it influence live migration? Thanks! diff --git a/src/config.h b/src/config.h index baca029..b3536f1 100644 index baca029..b3536f1 100644 --- a/src/config.h +++ b/src/config.h @@ -17,7 +17,7 @@ // Maximum number of map entries in the e820 map #define BUILD_MAX_E820 32 // Space to reserve in high-memory for tables -#define BUILD_MAX_HIGHTABLE (256*1024) +#define BUILD_MAX_HIGHTABLE (512*1024) diff --git a/src/malloc.c b/src/malloc.c index 3733855..6302525 100644 --- a/src/malloc.c +++ b/src/malloc.c @@ -549,13 +549,7 @@ malloc_prepboot(void) dprintf(1, "Space available for UMB: %x-%x, %x-%x\n" , RomEnd, base, info->range_start, info->range_end); - // Give back unused high ram. - info = alloc_find_lowest(&ZoneHigh); - if (info) { - u32 giveback = ALIGN_DOWN(info->range_end-info->range_start, PAGE_SIZE); - e820_add(info->range_start, giveback, E820_RAM); - dprintf(1, "Returned %d bytes of ZoneHigh\n", giveback); - } calcRamSize(); Regards, -Gonglei

1 0

[Question] can not boot oracle linux if increase the maximum size of permanent high memory area
by Xulei (Stone) 04 May '17

04 May '17

Hi，can you tell me what's the worst effect if i apply the following patch? In order to boot a BIG vm (with 4T mem, 255 vCPUs, 60 virtio-scsi disk...), i have to increase the BUILD_MAX_HIGHTABLE to 512KB. But, then i found i can not boot a specific VM anymore (oracle linux 6.7 64bits with kernel 3.8.13, 2G mem). It seems be related with the unused high ram given VM back by SeaBios (maybe also related to the memory align). I've tested if setting giveback memory 64KB alignment other than 4KB, SeaBios with 512KB BUILD_MAX_HIGHTABLE can boot the specific VM. But , i don't know the reason... As far as i know, if seabios does not give back the unused ZoneHigh memory seems will not trigger any problems but just waste only a little memory ,right? diff --git a/src/config.h b/src/config.h index baca029..b3536f1 100644 index baca029..b3536f1 100644 --- a/src/config.h +++ b/src/config.h @@ -17,7 +17,7 @@ // Maximum number of map entries in the e820 map #define BUILD_MAX_E820 32 // Space to reserve in high-memory for tables -#define BUILD_MAX_HIGHTABLE (256*1024) +#define BUILD_MAX_HIGHTABLE (512*1024) diff --git a/src/malloc.c b/src/malloc.c index 3733855..6302525 100644 --- a/src/malloc.c +++ b/src/malloc.c @@ -549,13 +549,7 @@ malloc_prepboot(void) dprintf(1, "Space available for UMB: %x-%x, %x-%x\n" , RomEnd, base, info->range_start, info->range_end); - // Give back unused high ram. - info = alloc_find_lowest(&ZoneHigh); - if (info) { - u32 giveback = ALIGN_DOWN(info->range_end-info->range_start, PAGE_SIZE); - e820_add(info->range_start, giveback, E820_RAM); - dprintf(1, "Returned %d bytes of ZoneHigh\n", giveback); - } calcRamSize(); ________________________________

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

SeaBIOS