Christian Hesse <list at eworm.de> on Tue, 2013/10/22 13:14:> Christian Hesse <list at eworm.de> on Tue, 2013/10/22 12:56: > > Gene Cumm <gene.cumm at gmail.com> on Tue, 2013/10/22 06:35: > > > On Tue, Oct 22, 2013 at 5:41 AM, Christian Hesse <list at eworm.de> wrote: > > > > Hello everybody, > > > > > > > > iPXE builds an ISO image by default. (The build system has to be > > > > updated for new paths and to integrate ldlinux.c32, but that's not my > > > > problem.) > > > > > > > > With the binary syslinux files from kernel.org booting fails with > > > > latest version 6.02: > > > > > > > > Loading ipxe.krn... ok > > > > Booting kernel failed: Invalid argument > > > > > > > > It does work with version 6.01, though. Any hints how to fix this > > > > regression? > > > > > > I presume this is using the precompiled binaries from kernel.org? > > > > Yes. > > > > > If > > > using the same ipxe.krn works on 6.01 and fails on 6.02, another thing > > > to try is to narrow it down to a -pre release. > > > > version 6.02-pre5 works, 6.02-pre6 does not. > > Struggled with git bisect, but finally succeeded: > > 8f470e7bfe75f6401f6c5432988c620b863ad274 is the first bad commit > commit 8f470e7bfe75f6401f6c5432988c620b863ad274 > Author: Matt Fleming <matt.fleming at intel.com> > Date: Tue Jul 16 15:15:54 2013 +0100 > > movebits: Add syslinux_memmap_find() > > Refactor the code for finding a suitable location for kernel > protected-mode and real-mode data. It's complicated enough that it > deserves to be separated into its own function. > > Signed-off-by: Matt Fleming <matt.fleming at intel.com> > > :040000 040000 a7f3dc3f6ac34bd09336eff39e2f3d2d1bd8c9b6 > 3e4cbdfc3cb9928c5c809fc4400623eefbc398f4 M com32If anybody tries to reproduce: I have uploaded an iPXE zImage that isolinux fails to boot: http://www.eworm.de/tmp/ipxe.lkrn Let be know if you need any more info. -- main(a){char*c=/* Schoene Gruesse */"B?IJj;MEH" "CX:;",b;for(a/* Chris get my mail address: */=0;b=c[a++];) putchar(b-1/(/* gcc -o sig sig.c && ./sig */b/42*2-3)*42);} -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 490 bytes Desc: not available URL: <http://www.zytor.com/pipermail/syslinux/attachments/20131024/dfb28814/attachment.sig>
Christian Hesse <list at eworm.de> on Thu, 2013/10/24 10:09:> Christian Hesse <list at eworm.de> on Tue, 2013/10/22 13:14: > > Christian Hesse <list at eworm.de> on Tue, 2013/10/22 12:56: > > > Gene Cumm <gene.cumm at gmail.com> on Tue, 2013/10/22 06:35: > > > > On Tue, Oct 22, 2013 at 5:41 AM, Christian Hesse <list at eworm.de> > > > > wrote: > > > > > Hello everybody, > > > > > > > > > > iPXE builds an ISO image by default. (The build system has to be > > > > > updated for new paths and to integrate ldlinux.c32, but that's not > > > > > my problem.) > > > > > > > > > > With the binary syslinux files from kernel.org booting fails with > > > > > latest version 6.02: > > > > > > > > > > Loading ipxe.krn... ok > > > > > Booting kernel failed: Invalid argument > > > > > > > > > > It does work with version 6.01, though. Any hints how to fix this > > > > > regression? > > > > > > > > I presume this is using the precompiled binaries from kernel.org? > > > > > > Yes. > > > > > > > If > > > > using the same ipxe.krn works on 6.01 and fails on 6.02, another thing > > > > to try is to narrow it down to a -pre release. > > > > > > version 6.02-pre5 works, 6.02-pre6 does not. > > > > Struggled with git bisect, but finally succeeded: > > > > 8f470e7bfe75f6401f6c5432988c620b863ad274 is the first bad commit > > commit 8f470e7bfe75f6401f6c5432988c620b863ad274 > > Author: Matt Fleming <matt.fleming at intel.com> > > Date: Tue Jul 16 15:15:54 2013 +0100 > > > > movebits: Add syslinux_memmap_find() > > > > Refactor the code for finding a suitable location for kernel > > protected-mode and real-mode data. It's complicated enough that it > > deserves to be separated into its own function. > > > > Signed-off-by: Matt Fleming <matt.fleming at intel.com> > > > > :040000 040000 a7f3dc3f6ac34bd09336eff39e2f3d2d1bd8c9b6 > > 3e4cbdfc3cb9928c5c809fc4400623eefbc398f4 M com32 > > If anybody tries to reproduce: I have uploaded an iPXE zImage that isolinux > fails to boot: > http://www.eworm.de/tmp/ipxe.lkrn > > Let be know if you need any more info.Is anybody working on this? Is there a bug tracker I should report this to so it does not get forgotten? Could not find one for syslinux. -- main(a){char*c=/* Schoene Gruesse */"B?IJj;MEH" "CX:;",b;for(a/* Chris get my mail address: */=0;b=c[a++];) putchar(b-1/(/* gcc -o sig sig.c && ./sig */b/42*2-3)*42);} -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 490 bytes Desc: not available URL: <http://www.zytor.com/pipermail/syslinux/attachments/20131111/f01c454a/attachment.sig>
On Mon, Nov 11, 2013 at 7:09 AM, Christian Hesse <list at eworm.de> wrote:> Christian Hesse <list at eworm.de> on Thu, 2013/10/24 10:09: >> Christian Hesse <list at eworm.de> on Tue, 2013/10/22 13:14: >> > Christian Hesse <list at eworm.de> on Tue, 2013/10/22 12:56: >> > > Gene Cumm <gene.cumm at gmail.com> on Tue, 2013/10/22 06:35: >> > > > On Tue, Oct 22, 2013 at 5:41 AM, Christian Hesse <list at eworm.de> >> > > > wrote: >> > > > > Hello everybody, >> > > > > >> > > > > iPXE builds an ISO image by default. (The build system has to be >> > > > > updated for new paths and to integrate ldlinux.c32, but that's not >> > > > > my problem.) >> > > > > >> > > > > With the binary syslinux files from kernel.org booting fails with >> > > > > latest version 6.02: >> > > > > >> > > > > Loading ipxe.krn... ok >> > > > > Booting kernel failed: Invalid argument >> > > > > >> > > > > It does work with version 6.01, though. Any hints how to fix this >> > > > > regression? >> > > > >> > > > I presume this is using the precompiled binaries from kernel.org? >> > > >> > > Yes. >> > > >> > > > If >> > > > using the same ipxe.krn works on 6.01 and fails on 6.02, another thing >> > > > to try is to narrow it down to a -pre release. >> > > >> > > version 6.02-pre5 works, 6.02-pre6 does not. >> > >> > Struggled with git bisect, but finally succeeded: >> > >> > 8f470e7bfe75f6401f6c5432988c620b863ad274 is the first bad commit >> > commit 8f470e7bfe75f6401f6c5432988c620b863ad274 >> > Author: Matt Fleming <matt.fleming at intel.com> >> > Date: Tue Jul 16 15:15:54 2013 +0100 >> > >> > movebits: Add syslinux_memmap_find() >> > >> > Refactor the code for finding a suitable location for kernel >> > protected-mode and real-mode data. It's complicated enough that it >> > deserves to be separated into its own function. >> > >> > Signed-off-by: Matt Fleming <matt.fleming at intel.com> >> > >> > :040000 040000 a7f3dc3f6ac34bd09336eff39e2f3d2d1bd8c9b6 >> > 3e4cbdfc3cb9928c5c809fc4400623eefbc398f4 M com32 >> >> If anybody tries to reproduce: I have uploaded an iPXE zImage that isolinux >> fails to boot: >> http://www.eworm.de/tmp/ipxe.lkrn >> >> Let be know if you need any more info. > > Is anybody working on this? > Is there a bug tracker I should report this to so it does not get forgotten? > Could not find one for syslinux.http://bugzilla.syslinux.org -- -Gene
On 11/11/2013 04:09 AM, Christian Hesse wrote:>> >> Let be know if you need any more info. > > Is anybody working on this? Is there a bug tracker I should report > this to so it does not get forgotten? Could not find one for > syslinux. >Hm... it is odd that 03dda0f1 etc. didn't fix this. Since this is extremely likely to be dependent on the memory map, this is likely to be very hard to determine without a debug trace, or at least the output of "meminfo.c32". -hpa
On 10/24/2013 01:09 AM, Christian Hesse wrote:>>> >>> version 6.02-pre5 works, 6.02-pre6 does not. >> >> Struggled with git bisect, but finally succeeded: >> >> 8f470e7bfe75f6401f6c5432988c620b863ad274 is the first bad commit >> commit 8f470e7bfe75f6401f6c5432988c620b863ad274 Author: Matt >> Fleming <matt.fleming at intel.com> Date: Tue Jul 16 15:15:54 2013 >> +0100 >> >> movebits: Add syslinux_memmap_find() >> >> Refactor the code for finding a suitable location for kernel >> protected-mode and real-mode data. It's complicated enough that >> it deserves to be separated into its own function. >> >> Signed-off-by: Matt Fleming <matt.fleming at intel.com> >> >> :040000 040000 a7f3dc3f6ac34bd09336eff39e2f3d2d1bd8c9b6 >> 3e4cbdfc3cb9928c5c809fc4400623eefbc398f4 M com32 > > If anybody tries to reproduce: I have uploaded an iPXE zImage that > isolinux fails to boot: http://www.eworm.de/tmp/ipxe.lkrn > > Let be know if you need any more info. >So this looks to be a zImage file (as opposed to a bzImage file), but it does have a substantial payload (0x59f50 bytes.) I am guessing that that is the source of the problems, but I'm still at a loss from staring at the code. If you could try turning on debugging output by having debug.c32 installed on the system and running: debug bios_boot_linux calc_cmdline_offset ... from the command line before trying to load ipxe.lkrn? I should try it out myself if I can find the time... -hpa
"H. Peter Anvin" <hpa at zytor.com> on Tue, 2013/12/03 20:26:> On 10/24/2013 01:09 AM, Christian Hesse wrote: > >>> > >>> version 6.02-pre5 works, 6.02-pre6 does not. > >> > >> Struggled with git bisect, but finally succeeded: > >> > >> 8f470e7bfe75f6401f6c5432988c620b863ad274 is the first bad commit > >> commit 8f470e7bfe75f6401f6c5432988c620b863ad274 Author: Matt > >> Fleming <matt.fleming at intel.com> Date: Tue Jul 16 15:15:54 2013 > >> +0100 > >> > >> movebits: Add syslinux_memmap_find() > >> > >> Refactor the code for finding a suitable location for kernel > >> protected-mode and real-mode data. It's complicated enough that > >> it deserves to be separated into its own function. > >> > >> Signed-off-by: Matt Fleming <matt.fleming at intel.com> > >> > >> :040000 040000 a7f3dc3f6ac34bd09336eff39e2f3d2d1bd8c9b6 > >> 3e4cbdfc3cb9928c5c809fc4400623eefbc398f4 M com32 > > > > If anybody tries to reproduce: I have uploaded an iPXE zImage that > > isolinux fails to boot: http://www.eworm.de/tmp/ipxe.lkrn > > > > Let be know if you need any more info. > > > > So this looks to be a zImage file (as opposed to a bzImage file), but > it does have a substantial payload (0x59f50 bytes.) > > I am guessing that that is the source of the problems, but I'm still > at a loss from staring at the code. If you could try turning on > debugging output by having debug.c32 installed on the system and running: > > debug bios_boot_linux calc_cmdline_offset > > ... from the command line before trying to load ipxe.lkrn? > > I should try it out myself if I can find the time...Sure. Here we go: boot: debug -e bios_boot_linux calc_cmdline_offset boot: ipxe.lkrn Loading ipxe.lkrn... ok cmdline_offset at 0x9f7e0 Initial memory map: Could not find location for protected-mode code Booting kernel failed: Invalid argument -- main(a){char*c=/* Schoene Gruesse */"B?IJj;MEH" "CX:;",b;for(a/* Chris get my mail address: */=0;b=c[a++];) putchar(b-1/(/* gcc -o sig sig.c && ./sig */b/42*2-3)*42);} -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 490 bytes Desc: not available URL: <http://www.zytor.com/pipermail/syslinux/attachments/20131204/f4b758c8/attachment.sig>