Re: Builtin microcode does nothing..

2016-06-03 Thread Borislav Petkov
On Sat, Jun 04, 2016 at 02:37:11AM +0200, Gabriel C wrote: > I tested your patches on both boxes on top 4.7-rc1 > and they work just fine for me. > > The test was done as before: > > CONFIG_FIRMWARE_IN_KERNEL=y > CONFIG_EXTRA_FIRMWARE="intel-ucode/" ( to match my CPU ) > CONFIG_BLK_DEV_

Re: Builtin microcode does nothing..

2016-06-03 Thread Gabriel C
On 03.06.2016 10:02, Borislav Petkov wrote: On Thu, May 26, 2016 at 02:46:06PM +0200, Borislav Petkov wrote: In the meantime, I've simplified the code even more and testing here with your .config looks good. I'll prepare another patchset next week and post it here in case you guys feel bored an

Re: Builtin microcode does nothing..

2016-06-03 Thread Borislav Petkov
On Thu, May 26, 2016 at 02:46:06PM +0200, Borislav Petkov wrote: > In the meantime, I've simplified the code even more and testing here > with your .config looks good. I'll prepare another patchset next week > and post it here in case you guys feel bored and want to test it too. Ok, here's a new v

Re: Builtin microcode does nothing..

2016-05-26 Thread Borislav Petkov
On Thu, May 26, 2016 at 01:52:22PM +0200, Gabriel C wrote: > With this patch ontop your your tip brach all is fine. > Tested on both server and both are up and running. Thanks. In the meantime, I've simplified the code even more and testing here with your .config looks good. I'll prepare anothe

Re: Builtin microcode does nothing..

2016-05-26 Thread Gabriel C
On 26.05.2016 12:03, Borislav Petkov wrote: On Thu, May 26, 2016 at 01:36:51AM +0200, Gabriel C wrote: Hangs on the second box too also.. Ok, please try the diff below ontop to see if it fixes your issue. Looks like I'd need to rewrite the figuring out where the microcode data is. Currently

Re: Builtin microcode does nothing..

2016-05-26 Thread Borislav Petkov
On Thu, May 26, 2016 at 01:36:51AM +0200, Gabriel C wrote: > Hangs on the second box too also.. Ok, please try the diff below ontop to see if it fixes your issue. Looks like I'd need to rewrite the figuring out where the microcode data is. Currently, it is ugly and error prone. Thanks. --- dif

Re: Builtin microcode does nothing..

2016-05-25 Thread Gabriel C
On 25.05.2016 23:50, Gabriel C wrote: I build an kernel on my oder box in a bit and let you know .. this server I tested on is sometimes strange Hangs on the second box too also..

Re: Builtin microcode does nothing..

2016-05-25 Thread Gabriel C
On 25.05.2016 23:38, Borislav Petkov wrote: On Wed, May 25, 2016 at 11:29:03PM +0200, Gabriel C wrote: initrd + buitl-in intel ucode won't boot here. It hangs on grub 'Loading initrd image'.. That's with my patches? Can I have the .config please so that I can reproduce? With initrd you mea

Re: Builtin microcode does nothing..

2016-05-25 Thread Borislav Petkov
On Wed, May 25, 2016 at 11:29:03PM +0200, Gabriel C wrote: > initrd + buitl-in intel ucode won't boot here. It hangs on grub 'Loading > initrd image'.. That's with my patches? Can I have the .config please so that I can reproduce? With initrd you mean, you supply an initrd with the micrcode with

Re: Builtin microcode does nothing..

2016-05-25 Thread Gabriel C
On 25.05.2016 11:31, Borislav Petkov wrote: On Sat, May 21, 2016 at 09:51:18AM +0200, Borislav Petkov wrote: I'll ping you once I'm done testing here. Ok, I've just uploaded a branch, it passes testing here. http://git.kernel.org/cgit/linux/kernel/git/bp/bp.git, branch tip-microcode I test

Re: Builtin microcode does nothing..

2016-05-25 Thread Borislav Petkov
On Wed, May 25, 2016 at 03:48:51PM +0200, Jim Bos wrote: > Works fine here: > > - no initrd.gz (simply all needed drivers builtin the main kernel) > - however, CONFIG_BLK_DEV_INITRD=y to keep option for initrd image > - intel ucode built-in kernel image > > It also survives suspend/resume S2RAM c

Re: Builtin microcode does nothing..

2016-05-25 Thread Jim Bos
On 05/25/2016 11:31 AM, Borislav Petkov wrote: > On Sat, May 21, 2016 at 09:51:18AM +0200, Borislav Petkov wrote: >> I'll ping you once I'm done testing here. > > Ok, I've just uploaded a branch, it passes testing here. > > http://git.kernel.org/cgit/linux/kernel/git/bp/bp.git, branch tip-microco

Re: Builtin microcode does nothing..

2016-05-25 Thread Borislav Petkov
On Sat, May 21, 2016 at 09:51:18AM +0200, Borislav Petkov wrote: > I'll ping you once I'm done testing here. Ok, I've just uploaded a branch, it passes testing here. http://git.kernel.org/cgit/linux/kernel/git/bp/bp.git, branch tip-microcode @Jim, I'd appreciate it if you ran it again, if you ge

Re: Builtin microcode does nothing..

2016-05-21 Thread Borislav Petkov
On Sat, May 21, 2016 at 04:59:15AM +0200, Gabriel C wrote: > Should this check not be the other way around ? Actually, I've changed it to this: /* try built-in microcode first */ if (load_builtin_intel_microcode(&cd)) /* * clear start as we might'v

Re: Builtin microcode does nothing..

2016-05-21 Thread Borislav Petkov
On Sat, May 21, 2016 at 02:20:56AM +0200, Gabriel C wrote: > I got to test an kernel without CONFIG_BLK_DEV_INITRD and this way > does work. Good, this confirms what I've been debugging here too. Thanks. -- Regards/Gruss, Boris. SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham

Re: Builtin microcode does nothing..

2016-05-20 Thread Gabriel C
On 21.05.2016 02:20, Gabriel C wrote: On 20.05.2016 12:08, Gabriel C wrote: Does it work when you disable CONFIG_BLK_DEV_INITRD in there? I can test this when I'm home since I need to change the config a bit. I got to test an kernel without CONFIG_BLK_DEV_INITRD and this way does work.

Re: Builtin microcode does nothing..

2016-05-20 Thread Gabriel C
On 20.05.2016 12:08, Gabriel C wrote: Does it work when you disable CONFIG_BLK_DEV_INITRD in there? I can test this when I'm home since I need to change the config a bit. I got to test an kernel without CONFIG_BLK_DEV_INITRD and this way does work. Regards, Gabriel C

Re: Builtin microcode does nothing..

2016-05-20 Thread Gabriel C
2016-05-20 12:08 GMT+02:00 Gabriel C : >> Does it work when you disable CONFIG_BLK_DEV_INITRD in there? > > I can test this whne I'm home since I need to change the config a bit. > >> I have a patchset which I'm testing right now which should correct some >> warts in builtin microcode, if you want

Re: Builtin microcode does nothing..

2016-05-20 Thread Borislav Petkov
On Fri, May 20, 2016 at 12:08:16PM +0200, Gabriel C wrote: > ( I think you forgot to CC me ) You're in To: From: Borislav Petkov To: Gabriel C Cc: LKML , Ingo Molnar , Thomas Gleixner Subject: Re: Builtin microcode does nothing.. Message-ID: <20160520083634.ga...@pd.tnic>

Re: Builtin microcode does nothing..

2016-05-20 Thread Gabriel C
> Does it work when you disable CONFIG_BLK_DEV_INITRD in there? I can test this whne I'm home since I need to change the config a bit. > I have a patchset which I'm testing right now which should correct some > warts in builtin microcode, if you want me to, I can push it out for you > to test too

Re: Builtin microcode does nothing..

2016-05-20 Thread Borislav Petkov
On Fri, May 20, 2016 at 02:34:34AM +0200, Gabriel C wrote: > With this kernel the early microcode loading does nothing nor the 'old > interface'. Does it work when you disable CONFIG_BLK_DEV_INITRD in there? I have a patchset which I'm testing right now which should correct some warts in builtin

Builtin microcode does nothing

2016-05-19 Thread Gabriel C
Hi guys , I tested kernel 4.6 with the built in microcode method , following Documentation/x86/early-microcode.txt. The kernel config[1] has : CONFIG_FIRMWARE_IN_KERNEL=y CONFIG_EXTRA_FIRMWARE="intel-ucode/06-1a-05" CONFIG_EXTRA_FIRMWARE_DIR="/lib/firmware" also CONFIG_MICROCODE=y CONFIG_MIC