SeaBIOS

seabios@seabios.org

10 participants
3770 discussions

Re: [SeaBIOS] [coreboot] [PATCH] SeaBIOS CD/DVD abbreviations
by Stefan Reinauer 11 Jun '10

11 Jun '10

On 6/9/10 5:43 PM, Andrew Morgan wrote: > Just a small point, the "ROM" bit isn't really needed. The fact that > CDs are read-only is pretty much irrelevant. No other type of disk > gets -ROM or -RAM stuck on the end of it. Also some CDs aren't read > only... Good point. Updated patch attached

2 1

[PATCH] seabios: remove iasl output file when error.
by Isaku Yamahata 11 Jun '10

11 Jun '10

Surprisingly iasl creates output file even when compilation error. So typing make after an error will succeed. This patch prevents it by removing the output file when error. And adds related dependencies to compile when .hex is missing. Signed-off-by: Isaku Yamahata <yamahata(a)valinux.co.jp> --- Makefile | 13 +++++++++---- 1 files changed, 9 insertions(+), 4 deletions(-) diff --git a/Makefile b/Makefile index 72d711d..d69c630 100644 --- a/Makefile +++ b/Makefile @@ -184,11 +184,16 @@ $(OUT)vgabios.bin: $(OUT)vgabios.bin.raw tools/buildrom.py $(Q)./tools/buildrom.py $< $@ ####### dsdt build rules -src/acpi-dsdt.hex: src/acpi-dsdt.dsl +$(OUT)%.dsl.i: src/%.dsl + $(Q)cpp -P $< > $@ + +src/%.hex: $(OUT)%.dsl.i @echo "Compiling DSDT" - $(Q)cpp -P $< > $(OUT)acpi-dsdt.dsl.i - $(Q)iasl -tc -p $@ $(OUT)acpi-dsdt.dsl.i - $(Q)rm $(OUT)acpi-dsdt.dsl.i + $(Q)iasl -tc -p $@ $< || (e=$$?; rm $@; exit $$e) + +src/acpi.c: src/acpi-dsdt.hex +$(OUT)ccode32flat.o: src/acpi.c +$(OUT)../src/acpi-dsdt.hex: src/acpi-dsdt.hex ####### Generic rules clean: -- 1.6.6.1

2 3

Re: [SeaBIOS] [PATCH] KVM vcpu hotplug seabios infrastructure
by Gleb Natapov 09 Jun '10

09 Jun '10

On Wed, Jun 09, 2010 at 09:20:07PM +0800, Liu, Jinsong wrote: > Which patch do you use? > I have send 2 patches before: > 1st patch is mis-send test patch which need manually build, > 2nd is updated patch to fix makefile issue, as attached email. > The patch you attached now compiles for me. Thanks. -- Gleb.

2 1

Re: [SeaBIOS] [PATCH] KVM vcpu hotplug seabios infrastructure
by Gleb Natapov 09 Jun '10

09 Jun '10

On Wed, Jun 09, 2010 at 05:38:41PM +0800, Liu, Jinsong wrote: > Gleb Natapov wrote: > >> > >> Currently we didn't test it with Windows 2008 datacenter. > >> > >> However, similar patch we also used for Xen (now already in Xen > >> upstream). > >> At Xen side, our QA team has tested with Windows 2008 datacenter. > >> For vcpu hot-add, it worked well, the new CPUs can be found on both > >> device manager and task manager. > >> For vcpu hot-remove, as you said, it cannot work since cpu > >> hot-remove is not supported by any Windows yet. > >> > >> At KVM side, we test the patch with linux 2.6.32, vcpu hot-add and > >> hot-remove both worked well. > >> If you need, we can co-work with our QA to test our KVM patch with > >> Windows 2008 datacenter. > >> > > It will be very helpful! Can you ask them to test the patch please? > > Gleb, > > Shaohui has done KVM vcpu hotplug test with Windows 2008 datacenter. > For vcpu hot-add, new CPUs can be found on device manager. > Thanks a lot for Shaohui's help! > Can't compile it: $ make Compiling whole program out/ccode.16.s Compiling (16bit) out/code16.o Compiling whole program out/ccode32flat.o src/acpi.c:274:28: error: acpi-ssdt-15.hex: No such file or directory src/acpi.c:275:32: error: acpi-ssdt-anycpu.hex: No such file or directory src/acpi.c: In function 'build_fadt': src/acpi.c:307: error: 'acpi_dsdt' undeclared (first use in this function) src/acpi.c:307: error: (Each undeclared identifier is reported only once src/acpi.c:307: error: for each function it appears in.) src/acpi.c: In function 'build_ssdt': src/acpi.c:435: error: 'acpi_ssdt_anycpu' undeclared (first use in this function) src/acpi.c:436: error: 'acpi_ssdt_15' undeclared (first use in this function) make: *** [out/ccode32flat.o] Error 1 -- Gleb.

2 1

Re: [SeaBIOS] [PATCH] KVM vcpu hotplug seabios infrastructure
by Gleb Natapov 09 Jun '10

09 Jun '10

On Mon, Jun 07, 2010 at 06:12:37PM +0800, Liu, Jinsong wrote: > Gleb Natapov wrote: > > On Tue, May 18, 2010 at 05:08:29PM +0800, Liu, Jinsong wrote: > >> Hi, Kevin, Gleb, Avi, and Yunhong > >> > >> I have updated vcpu hotplug seabios infrastructure, as attached. > >> Please kindly review it, thanks! > >> > >> -------------------------------------------- > >> This patch is to support KVM vcpu hotplug feature. > >> It ported from KVM bochs bios, and update some points to KVM seabios: > >> 1. build ssdt with 15/255 processor obj, which separately support > >> Win2000 and other oses; > >> 2. build madt accordingly, protect it as 'E820_NVS', and keep madt > >> checksum and lapic flag correctly when cpu hotplug; > >> 3. define Processor under \_SB instead of \_PR, so that it works > >> better with latest linux like 2.6.32 later; > >> 4. for Processor _MAT method, explicitly transfer it 'ToBuffer' so > >> that os will not confused with 'Integer' which will result in parse > >> failure; > >> 5. for Processor _STA method, use 0x0 instead of 0x9 for cpu > >> hot-remove, which matchs better with ACPI 4.0 spec, and works better > >> with latest linux; > >> 6. add Processor _EJ0 method, so that 'eject' appear under linux > >> /sys, and triggered by uevent to make cpu hot-remove; > >> ------------------------------------------- > >> > > Have you tested this patch with Windows 2008 datacenter edition? If > > yes were you able to insert vcpus so that Windows notice? (You need > > to test it with datacenter edition since others don't support cpu hot > > plug and cpu hot remove isn't supported by any Windows). > > Thanks Gleb for quick response :) Not really quick :) I am a bit busy currently with other things. > > Currently we didn't test it with Windows 2008 datacenter. > > However, similar patch we also used for Xen (now already in Xen upstream). > At Xen side, our QA team has tested with Windows 2008 datacenter. > For vcpu hot-add, it worked well, the new CPUs can be found on both device manager and task manager. > For vcpu hot-remove, as you said, it cannot work since cpu hot-remove is not supported by any Windows yet. > > At KVM side, we test the patch with linux 2.6.32, vcpu hot-add and hot-remove both worked well. > If you need, we can co-work with our QA to test our KVM patch with Windows 2008 datacenter. > It will be very helpful! Can you ask them to test the patch please? -- Gleb.

2 2

[PATCH] SeaBIOS CD/DVD abbreviations
by Stefan Reinauer 09 Jun '10

09 Jun '10

See Patch. -- coresystems GmbH • Brahmsstr. 16 • D-79104 Freiburg i. Br. Tel.: +49 761 7668825 • Fax: +49 761 7664613 Email: info(a)coresystems.de • http://www.coresystems.de/ Registergericht: Amtsgericht Freiburg • HRB 7656 Geschäftsführer: Stefan Reinauer • Ust-IdNr.: DE245674866

1 0

Re: [SeaBIOS] [PATCH] KVM vcpu hotplug seabios infrastructure
by Gleb Natapov 07 Jun '10

07 Jun '10

On Tue, May 18, 2010 at 05:08:29PM +0800, Liu, Jinsong wrote: > Hi, Kevin, Gleb, Avi, and Yunhong > > I have updated vcpu hotplug seabios infrastructure, as attached. Please kindly review it, thanks! > > -------------------------------------------- > This patch is to support KVM vcpu hotplug feature. > It ported from KVM bochs bios, and update some points to KVM seabios: > 1. build ssdt with 15/255 processor obj, which separately support Win2000 and other oses; > 2. build madt accordingly, protect it as 'E820_NVS', and keep madt checksum and lapic flag correctly when cpu hotplug; > 3. define Processor under \_SB instead of \_PR, so that it works better with latest linux like 2.6.32 later; > 4. for Processor _MAT method, explicitly transfer it 'ToBuffer' so that os will not confused with 'Integer' which will result in parse failure; > 5. for Processor _STA method, use 0x0 instead of 0x9 for cpu hot-remove, which matchs better with ACPI 4.0 spec, and works better with latest linux; > 6. add Processor _EJ0 method, so that 'eject' appear under linux /sys, and triggered by uevent to make cpu hot-remove; > ------------------------------------------- > Have you tested this patch with Windows 2008 datacenter edition? If yes were you able to insert vcpus so that Windows notice? (You need to test it with datacenter edition since others don't support cpu hot plug and cpu hot remove isn't supported by any Windows). -- Gleb.

2 1

Re: [SeaBIOS] [PATCH] KVM vcpu hotplug seabios infrastructure
by Kevin O'Connor 07 Jun '10

07 Jun '10

On Mon, May 24, 2010 at 08:18:37PM +0800, Liu, Jinsong wrote: > Gleb Natapov wrote: > > On Tue, May 18, 2010 at 05:08:29PM +0800, Liu, Jinsong wrote: > >> Hi, Kevin, Gleb, Avi, and Yunhong > >> > >> I have updated vcpu hotplug seabios infrastructure, as attached. > >> Please kindly review it, thanks! It's not clear to me what this patch tries to accomplish. Can you give an overview of what this patch does differently than the current code and why it is better? Also, it looks like your patch requires iasl to compile seabios. I'd prefer not to do that if possible. -Kevin

2 1

[PATCH] Allocate cdemu buffer in low mem instead of ebda.
by Kevin O'Connor 06 Jun '10

06 Jun '10

Using the low memory buffer gives more flexibility with the final location of the buffer. Don't allocate the buffer at all if no cdrom drives are present. --- src/biosvar.h | 2 -- src/cdrom.c | 33 ++++++++++++++++++--------------- src/post.c | 2 +- 3 files changed, 19 insertions(+), 18 deletions(-) diff --git a/src/biosvar.h b/src/biosvar.h index ea2d67d..dce35af 100644 --- a/src/biosvar.h +++ b/src/biosvar.h @@ -234,8 +234,6 @@ struct extended_bios_data_area_s { // Stack space available for code that needs it. u8 extra_stack[512] __aligned(8); - - u8 cdemu_buf[2048 * !!CONFIG_CDROM_EMU]; } PACKED; // The initial size and location of EBDA diff --git a/src/cdrom.c b/src/cdrom.c index f7af425..655ee00 100644 --- a/src/cdrom.c +++ b/src/cdrom.c @@ -17,6 +17,9 @@ * CD emulation ****************************************************************/ +struct drive_s *cdemu_drive_gf VAR16VISIBLE; +u8 *cdemu_buf_fl VAR16VISIBLE; + static int cdemu_read(struct disk_op_s *op) { @@ -30,12 +33,12 @@ cdemu_read(struct disk_op_s *op) int count = op->count; op->count = 0; - u8 *cdbuf_far = (void*)offsetof(struct extended_bios_data_area_s, cdemu_buf); + u8 *cdbuf_fl = GET_GLOBAL(cdemu_buf_fl); if (op->lba & 3) { // Partial read of first block. dop.count = 1; - dop.buf_fl = MAKE_FLATPTR(ebda_seg, cdbuf_far); + dop.buf_fl = cdbuf_fl; int ret = process_op(&dop); if (ret) return ret; @@ -43,10 +46,7 @@ cdemu_read(struct disk_op_s *op) if (thiscount > count) thiscount = count; count -= thiscount; - memcpy_far(FLATPTR_TO_SEG(op->buf_fl) - , (void*)FLATPTR_TO_OFFSET(op->buf_fl) - , ebda_seg, cdbuf_far + (op->lba & 3) * 512 - , thiscount * 512); + memcpy_fl(op->buf_fl, cdbuf_fl + (op->lba & 3) * 512, thiscount * 512); op->buf_fl += thiscount * 512; op->count += thiscount; dop.lba++; @@ -69,14 +69,12 @@ cdemu_read(struct disk_op_s *op) if (count) { // Partial read on last block. dop.count = 1; - dop.buf_fl = MAKE_FLATPTR(ebda_seg, cdbuf_far); + dop.buf_fl = cdbuf_fl; int ret = process_op(&dop); if (ret) return ret; u8 thiscount = count; - memcpy_far(FLATPTR_TO_SEG(op->buf_fl) - , (void*)FLATPTR_TO_OFFSET(op->buf_fl) - , ebda_seg, cdbuf_far, thiscount * 512); + memcpy_fl(op->buf_fl, cdbuf_fl, thiscount * 512); op->count += thiscount; } @@ -106,22 +104,27 @@ process_cdemu_op(struct disk_op_s *op) } } -struct drive_s *cdemu_drive_gf VAR16VISIBLE; - void cdemu_setup(void) { if (!CONFIG_CDROM_EMU) return; + cdemu_drive_gf = NULL; + cdemu_buf_fl = NULL; + if (!Drives.cdcount) + return; struct drive_s *drive_g = malloc_fseg(sizeof(*drive_g)); - if (! drive_g) { + u8 *buf = malloc_low(CDROM_SECTOR_SIZE); + if (!drive_g || !buf) { warn_noalloc(); - cdemu_drive_gf = NULL; + free(drive_g); + free(buf); return; } - memset(drive_g, 0, sizeof(*drive_g)); cdemu_drive_gf = drive_g; + cdemu_buf_fl = buf; + memset(drive_g, 0, sizeof(*drive_g)); drive_g->type = DTYPE_CDEMU; drive_g->blksize = DISK_SECTOR_SIZE; drive_g->sectors = (u64)-1; diff --git a/src/post.c b/src/post.c index 25535e2..fc7acc5 100644 --- a/src/post.c +++ b/src/post.c @@ -217,7 +217,6 @@ post(void) // Initialize internal tables boot_setup(); drive_setup(); - cdemu_setup(); // Start hardware initialization (if optionrom threading) if (CONFIG_THREADS && CONFIG_THREAD_OPTIONROMS) @@ -250,6 +249,7 @@ post(void) boot_prep(); // Finalize data structures before boot + cdemu_setup(); pmm_finalize(); malloc_finalize(); memmap_finalize(); -- 1.7.0.1

1 0

[PATCH] Introduce memcpy_fl - a memcpy on "flat" pointers.
by Kevin O'Connor 06 Jun '10

06 Jun '10

--- src/pmm.c | 20 +++++--------------- src/util.c | 11 +++++++++++ src/util.h | 1 + 3 files changed, 17 insertions(+), 15 deletions(-) diff --git a/src/pmm.c b/src/pmm.c index 228bc36..682be39 100644 --- a/src/pmm.c +++ b/src/pmm.c @@ -121,7 +121,6 @@ addSpace(struct zone_s *zone, void *start, void *end) tempdetail.datainfo.pprev = pprev; tempdetail.datainfo.data = tempdetail.datainfo.dataend = start; tempdetail.datainfo.allocend = end; - tempdetail.handle = PMM_DEFAULT_HANDLE; struct allocdetail_s *tempdetailp = MAKE_FLATPTR(GET_SEG(SS), &tempdetail); SET_PMMVAR(*pprev, &tempdetailp->datainfo); if (info) @@ -144,11 +143,8 @@ addSpace(struct zone_s *zone, void *start, void *end) } // Replace temp alloc space with final alloc space - SET_PMMVAR(detail->datainfo.next, tempdetail.datainfo.next); - SET_PMMVAR(detail->datainfo.pprev, tempdetail.datainfo.pprev); - SET_PMMVAR(detail->datainfo.data, tempdetail.datainfo.data); - SET_PMMVAR(detail->datainfo.dataend, tempdetail.datainfo.dataend); - SET_PMMVAR(detail->datainfo.allocend, tempdetail.datainfo.allocend); + memcpy_fl(&detail->datainfo, &tempdetailp->datainfo + , sizeof(detail->datainfo)); SET_PMMVAR(detail->handle, PMM_DEFAULT_HANDLE); SET_PMMVAR(*tempdetail.datainfo.pprev, &detail->datainfo); @@ -275,15 +271,9 @@ relocate_ebda(u32 newebda, u32 oldebda, u8 ebda_size) // EBDA isn't at end of ram - give up. return -1; - // Do copy - if (MODESEGMENT) - memcpy_far(FLATPTR_TO_SEG(newebda) - , (void*)FLATPTR_TO_OFFSET(newebda) - , FLATPTR_TO_SEG(oldebda) - , (void*)FLATPTR_TO_OFFSET(oldebda) - , ebda_size * 1024); - else - memmove((void*)newebda, (void*)oldebda, ebda_size * 1024); + // Do copy (this assumes memcpy copies forward - otherwise memmove + // is needed) + memcpy_fl((void*)newebda, (void*)oldebda, ebda_size * 1024); // Update indexes dprintf(1, "ebda moved from %x to %x\n", oldebda, newebda); diff --git a/src/util.c b/src/util.c index b2a22f7..2c22dfc 100644 --- a/src/util.c +++ b/src/util.c @@ -165,6 +165,17 @@ memcpy_far(u16 d_seg, void *d_far, u16 s_seg, const void *s_far, size_t len) : "cc", "memory"); } +inline void +memcpy_fl(void *d_fl, const void *s_fl, size_t len) +{ + if (MODESEGMENT) + memcpy_far(FLATPTR_TO_SEG(d_fl), (void*)FLATPTR_TO_OFFSET(d_fl) + , FLATPTR_TO_SEG(s_fl), (void*)FLATPTR_TO_OFFSET(s_fl) + , len); + else + memcpy(d_fl, s_fl, len); +} + void * #undef memcpy memcpy(void *d1, const void *s1, size_t len) diff --git a/src/util.h b/src/util.h index b475c42..7fd76bc 100644 --- a/src/util.h +++ b/src/util.h @@ -186,6 +186,7 @@ inline void memset16_far(u16 d_seg, void *d_far, u16 c, size_t len); void *memset(void *s, int c, size_t n); inline void memcpy_far(u16 d_seg, void *d_far , u16 s_seg, const void *s_far, size_t len); +void memcpy_fl(void *d_fl, const void *s_fl, size_t len); void *memcpy(void *d1, const void *s1, size_t len); #if MODESEGMENT == 0 #define memcpy __builtin_memcpy -- 1.7.0.1

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

SeaBIOS