All of lore.kernel.org
 help / color / mirror / Atom feed
From: geoff@infradead.org (Geoff Levand)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH v18 04/13] arm64/kexec: Add pr_debug output
Date: Thu, 16 Jun 2016 15:58:49 -0700	[thread overview]
Message-ID: <1466117929.21058.55.camel@infradead.org> (raw)
In-Reply-To: <57618CEB.6040807@arm.com>

On Wed, 2016-06-15 at 18:14 +0100, James Morse wrote:
> On 09/06/16 21:08, Geoff Levand wrote:
> > To aid in debugging kexec problems or when adding new functionality to kexec add
> > a new routine kexec_image_info() and several inline pr_debug statements.

> > diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
> > +/**
> > + * kexec_is_dtb - Helper routine to check the device tree header signature.
> > + */
> > +static bool kexec_is_dtb(const void *dtb)
> > +{
> > +> > 	> > __be32 magic;
> > +
> > +> > 	> > if (get_user(magic, (__be32 *)dtb))
> > +> > 	> > 	> > return false;
> > +
> 
> You pass this function 'kimage->segment[i].buf', this looks like the user space
> memory that contained the dtb when kexec-tools first ran to load the image.
> 
> This will work when you call it from machine_kexec_prepare(), but by the time we
> get to machine_kexec() that process is long gone, and this pointer should be
> considered junk.
> 
> I don't think its possible to find this information from machine_kexec(), as the
> DTB will be split up into page size chunks and scattered through memory.

That's correct.  I don't think it that important to print the
dtb segment, so I'll just remove this kexec_is_dtb() routine.

> > +
> >  void machine_kexec_cleanup(struct kimage *kimage)
> >  {
> >  > > 	> > /* Empty routine needed to avoid build errors. */
> > @@ -65,6 +107,8 @@ int machine_kexec_prepare(struct kimage *kimage)
> >  > > 	> > 	> > }
> >  > > 	> > }
> >  
> > +> > 	> > kexec_image_info(kimage);
> > +
> 
> You want to put this at the start of machine_kexec_prepare(), otherwise we may
> return from:
> > #ifdef CONFIG_HOTPLUG_CPU
> > 	> > /* any_cpu as we don't mind being preempted */
> > 	> > int any_cpu = raw_smp_processor_id();
> > 
> > 	> > if (cpu_ops[any_cpu]->cpu_die)
> > 	> > 	> > return 0;
> > #endif /* CONFIG_HOTPLUG_CPU */
> 
> This maybe-return-an-error block needs to be the last thing in the function.

OK.

> I'm not sure if the debug output is actually useful this early: kexec-tools
> prints out exactly the same information shortly after this function returns.

I thought it good to do it here, kexec-tools is not the only user
of kexec see https://github.com/antonblanchard/kexec-lite.

Thanks for checking.

-Geoff

WARNING: multiple messages have this Message-ID (diff)
From: Geoff Levand <geoff@infradead.org>
To: James Morse <james.morse@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>,
	marc.zyngier@arm.com, Catalin Marinas <catalin.marinas@arm.com>,
	Will Deacon <will.deacon@arm.com>,
	AKASHI Takahiro <takahiro.akashi@linaro.org>,
	kexec@lists.infradead.org, linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v18 04/13] arm64/kexec: Add pr_debug output
Date: Thu, 16 Jun 2016 15:58:49 -0700	[thread overview]
Message-ID: <1466117929.21058.55.camel@infradead.org> (raw)
In-Reply-To: <57618CEB.6040807@arm.com>

On Wed, 2016-06-15 at 18:14 +0100, James Morse wrote:
> On 09/06/16 21:08, Geoff Levand wrote:
> > To aid in debugging kexec problems or when adding new functionality to kexec add
> > a new routine kexec_image_info() and several inline pr_debug statements.

> > diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
> > +/**
> > + * kexec_is_dtb - Helper routine to check the device tree header signature.
> > + */
> > +static bool kexec_is_dtb(const void *dtb)
> > +{
> > +> > 	> > __be32 magic;
> > +
> > +> > 	> > if (get_user(magic, (__be32 *)dtb))
> > +> > 	> > 	> > return false;
> > +
> 
> You pass this function 'kimage->segment[i].buf', this looks like the user space
> memory that contained the dtb when kexec-tools first ran to load the image.
> 
> This will work when you call it from machine_kexec_prepare(), but by the time we
> get to machine_kexec() that process is long gone, and this pointer should be
> considered junk.
> 
> I don't think its possible to find this information from machine_kexec(), as the
> DTB will be split up into page size chunks and scattered through memory.

That's correct.  I don't think it that important to print the
dtb segment, so I'll just remove this kexec_is_dtb() routine.

> > +
> >  void machine_kexec_cleanup(struct kimage *kimage)
> >  {
> >  > > 	> > /* Empty routine needed to avoid build errors. */
> > @@ -65,6 +107,8 @@ int machine_kexec_prepare(struct kimage *kimage)
> >  > > 	> > 	> > }
> >  > > 	> > }
> >  
> > +> > 	> > kexec_image_info(kimage);
> > +
> 
> You want to put this at the start of machine_kexec_prepare(), otherwise we may
> return from:
> > #ifdef CONFIG_HOTPLUG_CPU
> > 	> > /* any_cpu as we don't mind being preempted */
> > 	> > int any_cpu = raw_smp_processor_id();
> > 
> > 	> > if (cpu_ops[any_cpu]->cpu_die)
> > 	> > 	> > return 0;
> > #endif /* CONFIG_HOTPLUG_CPU */
> 
> This maybe-return-an-error block needs to be the last thing in the function.

OK.

> I'm not sure if the debug output is actually useful this early: kexec-tools
> prints out exactly the same information shortly after this function returns.

I thought it good to do it here, kexec-tools is not the only user
of kexec see https://github.com/antonblanchard/kexec-lite.

Thanks for checking.

-Geoff


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

  reply	other threads:[~2016-06-16 22:58 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-06-09 20:08 [PATCH v18 00/13] arm64 kexec kernel patches Geoff Levand
2016-06-09 20:08 ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 06/13] arm64: kdump: reserve memory for crash dump kernel Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 03/13] arm64/kexec: Add core kexec support Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-15  8:14   ` AKASHI Takahiro
2016-06-15  8:14     ` AKASHI Takahiro
2016-06-15 17:10   ` James Morse
2016-06-15 17:10     ` James Morse
2016-06-16 22:41     ` Geoff Levand
2016-06-16 22:41       ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 02/13] arm64: Add cpus_are_stuck_in_kernel Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 05/13] arm64/kexec: Enable kexec in the arm64 defconfig Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 07/13] arm64: limit memory regions based on DT property, usable-memory Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 04/13] arm64/kexec: Add pr_debug output Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-15 17:14   ` James Morse
2016-06-15 17:14     ` James Morse
2016-06-16 22:58     ` Geoff Levand [this message]
2016-06-16 22:58       ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 08/13] arm64: kdump: implement machine_crash_shutdown() Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-14  7:13   ` AKASHI Takahiro
2016-06-14  7:13     ` AKASHI Takahiro
2016-06-14 21:56     ` Geoff Levand
2016-06-14 21:56       ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 01/13] arm64: Add back cpu reset routines Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 09/13] arm64: kdump: add kdump support Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 12/13] arm64: kdump: update a kernel doc Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 10/13] arm64: kdump: add VMCOREINFO for user-space coredump tools Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 13/13] Documentation: dt: usable-memory and elfcorehdr nodes for arm64 kexec Geoff Levand
2016-06-09 20:08   ` Geoff Levand
2016-06-09 20:08 ` [PATCH v18 11/13] arm64: kdump: enable kdump in the arm64 defconfig Geoff Levand
2016-06-09 20:08   ` Geoff Levand

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1466117929.21058.55.camel@infradead.org \
    --to=geoff@infradead.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.