October 2013 - coreboot - mail.coreboot.org

Help required to initialize coreboot as Seabios (floppy mechanism for DUET) payload
by Anthony Ross Oct. 19, 2013

Oct. 19, 2013

Hello Coreboot Team, Well Im very sorry that i forgot to mention in my previous email (Require help to boot coreboot with seabios' s floppy mechanism ) that the bootup of coreboot was initialized under qemu-kvm. I was able to perform this despite having no virtualization support. I have delayed a lot in responding and sorry for that too.Well I have begun a new thread and as suggested earlier by the coreboot team here is the output of `cbfstool build/coreboot.rom print`. coreboot.rom: 1024 kB, bootblocksize 848, romsize 1048576, offset 0x0 > alignment: 64 bytes > > Name Offset Type > Size > cmos_layout.bin 0x0 cmos_layout > 1160 > fallback/romstage 0x4c0 stage > 15390 > fallback/coreboot_ram 0x4140 stage > 43939 > fallback/payload 0xed40 payload > 63172 > config 0x1e440 raw > 3029 > floppyimg/MyFloppy.lzma 0x1f040 raw > 575728 > (empty) 0xab980 null > 344792 Not that this time I have booted coreboot with no hdd image & neither any virtual cd-rom image. Thanks Neo (attached coreboot and SeaBIOS log)

1 0

Require help to boot coreboot with seabios' s floppy mechanism
by Anthony Ross Oct. 19, 2013

Oct. 19, 2013

Well Im very sorry that i forgot to mention in my previous email (Require help to boot coreboot with seabios' s floppy mechanism ) that the bootup of coreboot was initialized under qemu-kvm. I was able to perform this despite having no virtualization support. I have delayed a lot in responding and sorry for that too.Well I have begun a new thread and as suggested earlier by the coreboot team here is the output of `cbfstool build/coreboot.rom print`. coreboot.rom: 1024 kB, bootblocksize 848, romsize 1048576, offset 0x0 > alignment: 64 bytes > > Name Offset Type > Size > cmos_layout.bin 0x0 cmos_layout > 1160 > fallback/romstage 0x4c0 stage > 15390 > fallback/coreboot_ram 0x4140 stage > 43939 > fallback/payload 0xed40 payload > 63172 > config 0x1e440 raw > 3029 > floppyimg/MyFloppy.lzma 0x1f040 raw > 575728 > (empty) 0xab980 null > 344792 (attached coreboot and SeaBIOS log)

1 0

EHCI Debug Gadget
by John Lewis Oct. 17, 2013

Oct. 17, 2013

Hey Guys, I'm trying to setup a Beaglebone Black as the subject. I've followed the instructions and I get lots of this when I try to make it do it's thang: [ 3142.831914] CAUTION: musb: Babble Interrupt Occurred [ 3143.107609] gadget: setup: failure req 9 v 1 [ 3143.289034] CAUTION: musb: Babble Interrupt Occurred [ 3143.289088] gadget: disconnected [ 3143.564615] gadget: setup: failure req 9 v 1 [ 3143.746013] CAUTION: musb: Babble Interrupt Occurred [ 3143.746058] gadget: disconnected [ 3144.021599] gadget: setup: failure req 9 v 1 [ 3144.264994] CAUTION: musb: Babble Interrupt Occurred [ 3144.265038] gadget: disconnected [ 3144.540802] gadget: setup: failure req 9 v 1 [ 3144.722247] CAUTION: musb: Babble Interrupt Occurred [ 3144.722304] gadget: disconnected [ 3144.997944] gadget: setup: failure req 9 v 1 [ 3145.179357] CAUTION: musb: Babble Interrupt Occurred [ 3145.179415] gadget: disconnected Any pointers? Web search comes up fruitless except for references to the kernel source code which I don't understand. ;) Thanks, John.

2 3

Re: [coreboot] The OS of open source solution
by Patrick Georgi Oct. 17, 2013

Oct. 17, 2013

Am 17.10.2013 09:38, schrieb Stojsavljevic, Zoran: > For Windows 7, it has two boot modes: legacy 32-bit boot mode, and > UEFI 64-bit boot mode. The former/legacy boot mode boots from any > legacy BIOS (SeaBIOS, AMI, Intel, etc.), and the latter one requires > few UEFI run-time services. > They say that Windows 8 only boots from > UEFI, but it might be possible to boot with SeaBIOS using 32-bit > Windows is becoming possible. Run-time services will not disappear > during or after boot, but all other boot firmware will. Both 32bit and 64bit Windows, both 7 and 8, are supposed to be bootable with PC BIOS. It's just that logo requirements mandate UEFI for Windows 8, but that says nothing about its capability to boot with SeaBIOS. When booting with UEFI, 64bit is required, since they only interface with the 64bit version of the UEFI APIs. However you can't boot a BIOS-installed Windows with UEFI and vice-versa. So if you installed an UEFI Windows, SeaBIOS won't be able to run it. Solution: Do fresh Windows installs from BIOS (UEFI+CSM is enough), then use these for your coreboot experiments. If you're creating new installations, consider obtaining "checked" Windows builds (available through MSDN channels). Those are much more useful for debugging, in particular when it comes to ACPI issues (with one caveat, see http://www.coreboot.org/ACPI#Using_checked_builds) Patrick

2 1

Re: [coreboot] Require help to boot coreboot with seabios' s floppy mechanism
by Paul Menzel Oct. 17, 2013

Oct. 17, 2013

Dear Neo, Am Donnerstag, den 17.10.2013, 12:32 +0530 schrieb Anthony Ross: > I have successfully build coreboot with seabios as payload. I can well boot > a HDD Image & so also a virtual cd-rom image i.e. a Ubuntu 12.04 ISO. awesome. What board and coreboot and SeaBIOS versions do you use? > However I have further got on building the DUET Package. Could you please reply to the other thread »Help required for compiling EDK2 ' s CorebootPkg« from October 3rd [1], what you did to get it to build. > Since in case of coreboot it can' t be added as a true tianocore > payload I switched to seabios' s floppy mechanism to load DUET. > Well after adding the lzma compressed floppy image i can well embed it in > coreboot.rom flash image, but the main problem is that it does not appear > in the seabios' s boot options. When i press F12 it only shows a virtual > DVD drive ATA 1 & thats all.I have well followed the > coreboot/tianocore <http://www.coreboot.org/TianoCore> build guide for > DUET and embedding it. Please provide us the output of `cbfstool build/coreboot.rom print`. Also please build SeaBIOS with debug level 8 and attach the coreboot and SeaBIOS log (use serial console, CBMEM or USB debug dongle). Thanks, Paul [1] http://www.coreboot.org/pipermail/coreboot/2013-October/thread.html#76416

1 1

Re: [coreboot] The OS of open source solution
by 汤科峰 Oct. 17, 2013

Oct. 17, 2013

Hi all, Now I can install&boot OpenSuse with "coreboot(intel fsp)+seabios". For win7 & win8 ,I think that tianocore as payload is a better way. Intel Fsp provide HOB pointer,we maybe need to hand off this to DXE pharse and then do mainboard init ..... Hi Zoran, What do you think of this idea? Thanks, TankTang At 2013-10-16 20:47:38,"Stojsavljevic, Zoran" <zoran.stojsavljevic(a)intel.com> wrote: >Hello to everyone, > >On Monday I was racing with the time, so I just sent email to acknowledge that I was able to bring FSP concept to 3.0.13 kernel (and disappeared few minutes later). Day and a half I was busy with completely another job. > >First, I would like to thank to Patrick (Georgi) for unselfish help Patrick provided to me. Thank you, Patrick! > >Second, David, my answers are down below regarding Win7 and Win8. Today I have started testing more than kernel 3.0.13 (which I have adopted to Bare Metal OS, since I am embedded guy). I found that very few kernels so far are booting. But 3.0.13 without flaw, constantly coming up! > >I tried to do more impossible things, knowing that they will not work for now. These are to boot Win7 and Win8. I had Win7 HDD for IVB, and installed one HDD for Win8 using IVB CRB with EFI BIOS. Win8 seamlessly boots with IVB CRB with EFI BIOS. Knowing that Win7 requires EFI or UEFI BIOS (mandatory using EFI/UEFI BIOS drivers), I did not expect Win7 to boot. The pleasant surprise was that Win7 bootloader tried to boot Win7, concluded that Win7 is broken/damaged (NO EFI drivers), and safely redirected me back to Win7 bootloader VGA screen asking me for the options: > Launch Startup Repair (recommended) > Starts Windows Normally > >With Win8, I had different story in my mind. I assumed (exchanging some thoughts internally with some designers) that Win8 is "dual", so it'll go to discover EFI/UEFI, but if it fails, it has different mechanisms to boot without EFI/UEFI. I tried to boot 3 times. All three times the blue Win8 flag/logo appeared with the message: Preparing Automatic Repair. And froze. > >Log files attached for both cases. > >Any thoughts/ideas/theories? > >Thank you, >Zoran >_______ >Most of The Time you should be "intel inside" to be capable to think "out of the box". >_______ > >From: David Hendricks [mailto:david.hendricks@gmail.com] >Sent: Monday, October 14, 2013 10:39 PM >To: Stojsavljevic, Zoran >Cc: ron minnich; jstkf2012(a)126.com; David Hendricks; coreboot >Subject: Re: [coreboot] The OS of open source solution > >On Mon, Oct 14, 2013 at 6:48 AM, Stojsavljevic, Zoran <zoran.stojsavljevic(a)intel.com> wrote: >Well... > >Few minutes ago, I finally was able to do the impossible: To boot in this sequence: IVB FSP -> Coreboot -> SeaBIOS -> GRUB 0.97 (SuSE SLES 11 SP1) -> Linux 3.0.13 with Bare Metal capabilities. > >Congratulations! > >Do you have a copy of Windows 7 to try to install via CD-ROM? Maybe your method can help TankTang. > >Thank you for the great support! >Zoran >_______ >Most of The Time you should be "intel inside" to be capable to think "out of the box". >-----Original Message----- >From: ron minnich [mailto:rminnich@gmail.com] >Sent: Wednesday, October 09, 2013 8:02 AM >To: jstkf2012(a)126.com >Cc: Stojsavljevic, Zoran; David Hendricks; coreboot >Subject: Re: Re: [coreboot] The OS of open source solution > >well, we are all interested in how to incorporate fsp into coreboot. >And you have an FSP binary. It seems to me we have a place to start. > >ron > >On Tue, Oct 8, 2013 at 10:02 PM, jstkf2012(a)126.com <jstkf2012(a)126.com> wrote: >> Hi Ron, >> I am sorry , I am just a freshman in CoreBoot. >> Do you interest in how to incorporate the Fsp into Coreboot?But the >> source code and the mechanism is not writed and created by myself. >> >> Thanks, >> Tank >> From: ron minnich >> Date: 2013-10-09 10:07 >> To: jstkf2012(a)126.com >> CC: Stojsavljevic, Zoran; David Hendricks; coreboot >> Subject: Re: Re: [coreboot] The OS of open source solution On Tue, Oct >> 8, 2013 at 6:20 PM, jstkf2012(a)126.com <jstkf2012(a)126.com> wrote: >>> Hi Ron >>> >>> I doesn't not drop fsp in to replace the mrc binary. >> >> We are very interested in what you are doing and if we can help in >> some way, please let us know. >> >> Thanks! >> >> ron > >-- >coreboot mailing list: coreboot(a)coreboot.org >http://www.coreboot.org/mailman/listinfo/coreboot > >Intel GmbH >Dornacher Strasse 1 >85622 Feldkirchen/Muenchen, Deutschland >Sitz der Gesellschaft: Feldkirchen bei Muenchen >Geschaeftsfuehrer: Christian Lamprechter, Hannes Schwaderer, Douglas Lusk >Registergericht: Muenchen HRB 47456 >Ust.-IdNr./VAT Registration No.: DE129385895 >Citibank Frankfurt a.M. (BLZ 502 109 00) 600119052

3 2

Re: [coreboot] Require help to boot coreboot with seabios' s floppy mechanism
by Andrew Wu Oct. 17, 2013

Oct. 17, 2013

Hi, First you should make sure you enabled "Floppy images in CBFS" option in SeaBIOS. It is in menuconfig -> General Features -> Floppy images in CBFS. You can also check your .config file for FLASH_FLOPPY setting. SeaBIOS only accepts floppy image file size that is the same as a physical floppy driver. You can check SeaBIOS RAM disk source code: http://code.coreboot.org/p/seabios/source/tree/master/src/hw/ramdisk.c It calls to find_floppy_type. It is in: http://code.coreboot.org/p/seabios/source/tree/master/src/hw/floppy.c You can see find_floppy_type will lookup floppy driver size in FloppyInfo array. They are supported floppy image file sizes. (360KB, 1.2MB, etc) I am not sure if SeaBIOS support compressed floppy image file. You can try to use an uncompressed floppy image file. 2013/10/17 Anthony Ross <anthonyross230(a)gmail.com>: > Hi there: > > I have successfully build coreboot with seabios as payload. I can well boot > a HDD Image & so also a virtual cd-rom image i.e. a Ubuntu 12.04 ISO. > However I have futher got on building the DUET Package. Since in case of > coreboot it can' t be added as a true tianocore payload I switched to > seabios' s floppy mechanism to load DUET. > Well after adding the lzma compressed floppy image i can well embed it in > coreboot.rom flash image, but the main problem is that it does not appear in > the seabios' s boot options. When i press F12 it only shows a virtual DVD > drive ATA 1 & thats all.I have well followed the coreboot/tianocore build > guide for DUET and embedding it. > > > Any help would be greatly appreciated.. > > > Neo > > -- > coreboot mailing list: coreboot(a)coreboot.org > http://www.coreboot.org/mailman/listinfo/coreboot

1 0

Require help to boot coreboot with seabios' s floppy mechanism
by Anthony Ross Oct. 17, 2013

Oct. 17, 2013

Hi there: I have successfully build coreboot with seabios as payload. I can well boot a HDD Image & so also a virtual cd-rom image i.e. a Ubuntu 12.04 ISO. However I have futher got on building the DUET Package. Since in case of coreboot it can' t be added as a true tianocore payload I switched to seabios' s floppy mechanism to load DUET. Well after adding the lzma compressed floppy image i can well embed it in coreboot.rom flash image, but the main problem is that it does not appear in the seabios' s boot options. When i press F12 it only shows a virtual DVD drive ATA 1 & thats all.I have well followed the coreboot/tianocore<http://www.coreboot.org/TianoCore>build guide for DUET and embedding it. Any help would be greatly appreciated.. Neo

1 0

cpu/amd/car/post_cache_as_ram.c: Build problems with GCC 4.8.x: `error: bp cannot be used in asm here`
by Paul Menzel Oct. 16, 2013

Oct. 16, 2013

Dear coreboot folks, I submitted a bug report to GCC about the build failures we got with GCC 4.8.x¹. Björn Busse noticed that the ASUS M2V-MX SE did not build anymore. (This was probably not detected by Jenkins because its coreboot toolchain has to be updated manually.) The code is in `src/cpu/amd/car/post_cache_as_ram.c` [1] __asm__ volatile ( /* set new esp */ /* before CONFIG_RAMBASE */ "subl %0, %%esp\n\t" ::"a"( (CONFIG_DCACHE_RAM_BASE + CONFIG_DCACHE_RAM_SIZE)- (CONFIG_RAMTOP) ) /* discard all registers (eax is used for %0), so gcc redoes everything after the stack is moved */ : "cc", "memory", "%ebx", "%ecx", "%edx", "%esi", "%edi", "%ebp" ); printf("Done.\n"); and gcc throws the following error. $ make V=1 […] CC romstage.inc /src/coreboot/util/crossgcc/xgcc/bin/i386-elf-gcc -Wa,--divide -fno-stack-protector -Wl,--build-id=none -MMD -Isrc -Isrc/include -Ibuild -Isrc/arch/x86/include -Isrc/device/oprom/include -include src/include/kconfig.h -Os -pipe -g -nostdinc -nostdlib -Wall -Wundef -Wstrict-prototypes -Wmissing-prototypes -Wwrite-strings -Wredundant-decls -Wno-trigraphs -Wstrict-aliasing -Wshadow -Werror -fno-common -ffreestanding -fno-builtin -fomit-frame-pointer -D__PRE_RAM__ -Isrc -I. -Ibuild -c -S src/mainboard/asus/m2v-mx_se/romstage.c -o build/mainboard/asus/m2v-mx_se/romstage.pre.inc src/mainboard/asus/m2v-mx_se/romstage.c: In function 'pci_locate_device': src/mainboard/asus/m2v-mx_se/romstage.c:196:1: internal compiler error: Illegal instruction Please submit a full bug report, with preprocessed source if appropriate. See <http://gcc.gnu.org/bugs.html> for instructions. make: *** [build/mainboard/asus/m2v-mx_se/romstage.pre.inc] Error 1 The verbose error is `error: bp cannot be used in asm here`. The offending code lies in and therefore the patch was reverted by Patrick Georgi in commit c8883262 (buildgcc: Downgrade to gcc 4.7.3, handle armv7-a) [2]. My ticket 58758 [3] was closed right away as invalid by Andrew Pinski and he linked to bug 11807 [4]. Furthermore he replied the following about the code used in coreboot. This code really should be written in pure asm rather than doing an inline-asm as GCC can do many different optimizations more than just rereading them from the stack. So it looks like GCC is not going to be changed/fixed and, if we wanted to update coreboot’s toolchain, we would need to fix our code. Thanks, Paul ¹ Sometimes it fails for me with 4.7.3 too. I do not know why, but it might be related to `.xcompile` not having some flags it should have. This is another problem though. [1] http://review.coreboot.org/3930 [2] http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58758 [3] http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11807

2 1

The OS of open source solution
by Free_QP＠hotmail.com Oct. 16, 2013

Oct. 16, 2013

Hi Maillinglist, I’m now trying to boot OS with open source solution (Coreboot + SeaBIOS), but have some questions. For example, 1.Which OS can be supported by this kind of solution? For example, is Win7 or Win8 supported? 2.How to build/create or customize the OS images? Thanks， Free_QP

2 1