Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1300569 > unrolled thread

Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc() for iomem search

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc()  for iomem search Dave Young <dyoung@redhat.com> - 2016-01-04 10:30 +0100
    Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc()  for iomem search Borislav Petkov <bp@alien8.de> - 2016-01-04 13:30 +0100
      Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc()  for iomem search Toshi Kani <toshi.kani@hpe.com> - 2016-01-04 19:00 +0100
        Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc()  for iomem search Toshi Kani <toshi.kani@hpe.com> - 2016-01-04 20:50 +0100
        Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc()  for iomem search Borislav Petkov <bp@alien8.de> - 2016-01-04 20:50 +0100
      Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc()  for iomem search Dave Young <dyoung@redhat.com> - 2016-01-05 06:30 +0100

#1300569 — Re: [PATCH v2 14/16] x86, nvdimm, kexec: Use walk_iomem_res_desc() for iomem search

Hi, Boris

On 12/27/15 at 11:24am, Borislav Petkov wrote:
> On Sun, Dec 27, 2015 at 10:12:57AM +0800, Minfei Huang wrote:
> > You can refer the below link that you may get a clue about GART. This is
> > the fisrt time kexec-tools tried to support to ignore GART region in 2nd
> > kernel.
> > 
> > http://lists.infradead.org/pipermail/kexec/2008-December/003096.html
> 
> So theoretically we could export that IORES_DESC* enum in an uapi header and
> move kexec-tools to use that.
> 
> However, I'm fuzzy on how exactly the whole compatibility thing is done
> with kexec-tools and the kernel. We probably would have to support
> newer kexec-tools on an older kernel and vice versa so I'd guess we
> should mark walk_iomem_res() deprecated so that people are encouraged to
> upgrade to newer kexec-tools...

Replied to Toshi old kernel will export the "GART" region for amd cards.
So for old kernel and new kexec-tools we will have problem.

I think add the GART desc for compitibility purpose is doable, no?

Thanks
Dave
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [next] | [standalone]

#1300698

On Mon, Jan 04, 2016 at 05:29:37PM +0800, Dave Young wrote:
> Replied to Toshi old kernel will export the "GART" region for amd cards.
> So for old kernel and new kexec-tools we will have problem.
> 
> I think add the GART desc for compitibility purpose is doable, no?

Just read your other mails too. If I see it correctly, there's only one
place which has "GART":

$ git grep -e \"GART\"
arch/x86/kernel/crash.c:235:    walk_iomem_res("GART", IORESOURCE_MEM, 0, -1,

So crash.c only excludes this region but the kernel doesn't create it.
Right?

So we can kill that walk_iomem_res(), as you say. Which would be even
nicer...

-- 
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1300905

On Mon, 2016-01-04 at 13:26 +0100, Borislav Petkov wrote:
> On Mon, Jan 04, 2016 at 05:29:37PM +0800, Dave Young wrote:
> > Replied to Toshi old kernel will export the "GART" region for amd
> > cards.
> > So for old kernel and new kexec-tools we will have problem.
> > 
> > I think add the GART desc for compitibility purpose is doable, no?
> 
> Just read your other mails too. If I see it correctly, there's only one
> place which has "GART":
> 
> $ git grep -e \"GART\"
> arch/x86/kernel/crash.c:235:    walk_iomem_res("GART", IORESOURCE_MEM, 0,
> -1,
> 
> So crash.c only excludes this region but the kernel doesn't create it.
> Right?
> 
> So we can kill that walk_iomem_res(), as you say. Which would be even
> nicer...

Agreed.  As Dave suggested in the other thread, we can simply remove the
walk_iomem_res("GART",) call in crash.c.  

With this change, there will be no caller to walk_iomem_res().  Should we
remove walk_iomem_res() altogether, or keep it for now as a deprecated func
with the checkpatch check? 

Thanks,
-Toshi

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1300996

On Mon, 2016-01-04 at 20:41 +0100, Borislav Petkov wrote:
> On Mon, Jan 04, 2016 at 10:57:40AM -0700, Toshi Kani wrote:
> > With this change, there will be no caller to walk_iomem_res().  Should 
> > we remove walk_iomem_res() altogether, or keep it for now as a 
> > deprecated func with the checkpatch check?
> 
> Yes, kill it on the spot so that people don't get crazy ideas.

Will do.  

Thanks!
-Toshi

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1301000

On Mon, Jan 04, 2016 at 10:57:40AM -0700, Toshi Kani wrote:
> With this change, there will be no caller to walk_iomem_res().  Should we
> remove walk_iomem_res() altogether, or keep it for now as a deprecated func
> with the checkpatch check?

Yes, kill it on the spot so that people don't get crazy ideas.

Thanks!

-- 
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1301277

On 01/04/16 at 01:26pm, Borislav Petkov wrote:
> On Mon, Jan 04, 2016 at 05:29:37PM +0800, Dave Young wrote:
> > Replied to Toshi old kernel will export the "GART" region for amd cards.
> > So for old kernel and new kexec-tools we will have problem.
> > 
> > I think add the GART desc for compitibility purpose is doable, no?
> 
> Just read your other mails too. If I see it correctly, there's only one
> place which has "GART":
> 
> $ git grep -e \"GART\"
> arch/x86/kernel/crash.c:235:    walk_iomem_res("GART", IORESOURCE_MEM, 0, -1,
> 
> So crash.c only excludes this region but the kernel doesn't create it.
> Right?

Right.

> 
> So we can kill that walk_iomem_res(), as you say. Which would be even
> nicer...

Yes, I think it is ok to kill walk_iomem_res()

Thanks
Dave
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web