All of lore.kernel.org
 help / color / mirror / Atom feed
From: Sasha Levin <sasha.levin@oracle.com>
To: David Rientjes <rientjes@google.com>
Cc: linux-mm@kvack.org, akpm@linux-foundation.org,
	linux-kernel@vger.kernel.org, kirill@shutemov.name
Subject: Re: [PATCH 05/11] mm: debug: dump page into a string rather than directly on screen
Date: Wed, 01 Jul 2015 15:21:37 -0400	[thread overview]
Message-ID: <55943DC1.6010209@oracle.com> (raw)
In-Reply-To: <alpine.DEB.2.10.1506301627030.5359@chino.kir.corp.google.com>

On 06/30/2015 07:35 PM, David Rientjes wrote:
> I don't know how others feel, but this looks strange to me and seems like 
> it's only a result of how we must now dump page information 
> (dump_page(page) is no longer available, we must do pr_alert("%pZp", 
> page)).
> 
> Since we're relying on print formats, this would arguably be better as
> 
> 	pr_alert("Not movable balloon page:\n");
> 	pr_alert("%pZp", page);
> 
> to avoid introducing newlines into potentially lengthy messages that need 
> a specified loglevel like you've done above.
> 
> But that's not much different than the existing dump_page() 
> implementation.
> 
> So for this to be worth it, it seems like we'd need a compelling usecase 
> for something like pr_alert("%pZp %pZv", page, vma) and I'm not sure we're 
> ever actually going to see that.  I would argue that
> 
> 	dump_page(page);
> 	dump_vma(vma);
> 
> would be simpler in such circumstances.

I think we can find usecases where we want to dump more information than what's
contained in just one page/vma/mm struct. Things like the following from mm/gup.c:

	VM_BUG_ON_PAGE(compound_head(page) != head, page);

Where seeing 'head' would be interesting as well.

Or for VMAs, from include/linux/rmap.h:

	VM_BUG_ON_VMA(vma->anon_vma != next->anon_vma, vma);

Would it be interesting to see both vma, and next? Probably.

Or opportunities to add information from other variables, such as in:

	VM_BUG_ON_PAGE(stable_node->kpfn != page_to_pfn(oldpage), oldpage);

Is stable_node->kpfn interesting? Might be.


We *could* go ahead and open code all of that, but that's not happening, It's not
intuitive and people just slap VM_BUG_ON()s and hope they can figure it out when
those VM_BUG_ON()s happen.

Are there any pieces of code that open code what you suggested?


Thanks,
Sasha

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

WARNING: multiple messages have this Message-ID (diff)
From: Sasha Levin <sasha.levin@oracle.com>
To: David Rientjes <rientjes@google.com>
Cc: linux-mm@kvack.org, akpm@linux-foundation.org,
	linux-kernel@vger.kernel.org, kirill@shutemov.name
Subject: Re: [PATCH 05/11] mm: debug: dump page into a string rather than directly on screen
Date: Wed, 01 Jul 2015 15:21:37 -0400	[thread overview]
Message-ID: <55943DC1.6010209@oracle.com> (raw)
In-Reply-To: <alpine.DEB.2.10.1506301627030.5359@chino.kir.corp.google.com>

On 06/30/2015 07:35 PM, David Rientjes wrote:
> I don't know how others feel, but this looks strange to me and seems like 
> it's only a result of how we must now dump page information 
> (dump_page(page) is no longer available, we must do pr_alert("%pZp", 
> page)).
> 
> Since we're relying on print formats, this would arguably be better as
> 
> 	pr_alert("Not movable balloon page:\n");
> 	pr_alert("%pZp", page);
> 
> to avoid introducing newlines into potentially lengthy messages that need 
> a specified loglevel like you've done above.
> 
> But that's not much different than the existing dump_page() 
> implementation.
> 
> So for this to be worth it, it seems like we'd need a compelling usecase 
> for something like pr_alert("%pZp %pZv", page, vma) and I'm not sure we're 
> ever actually going to see that.  I would argue that
> 
> 	dump_page(page);
> 	dump_vma(vma);
> 
> would be simpler in such circumstances.

I think we can find usecases where we want to dump more information than what's
contained in just one page/vma/mm struct. Things like the following from mm/gup.c:

	VM_BUG_ON_PAGE(compound_head(page) != head, page);

Where seeing 'head' would be interesting as well.

Or for VMAs, from include/linux/rmap.h:

	VM_BUG_ON_VMA(vma->anon_vma != next->anon_vma, vma);

Would it be interesting to see both vma, and next? Probably.

Or opportunities to add information from other variables, such as in:

	VM_BUG_ON_PAGE(stable_node->kpfn != page_to_pfn(oldpage), oldpage);

Is stable_node->kpfn interesting? Might be.


We *could* go ahead and open code all of that, but that's not happening, It's not
intuitive and people just slap VM_BUG_ON()s and hope they can figure it out when
those VM_BUG_ON()s happen.

Are there any pieces of code that open code what you suggested?


Thanks,
Sasha

  parent reply	other threads:[~2015-07-01 19:30 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-05-14 17:10 [PATCH 00/11] mm: debug: formatting memory management structs Sasha Levin
2015-05-14 17:10 ` Sasha Levin
2015-05-14 17:10 ` [PATCH 01/11] mm: debug: format flags in a buffer Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 02/11] mm: debug: deal with a new family of MM pointers Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 03/11] mm: debug: dump VMA into a string rather than directly on screen Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 04/11] mm: debug: dump struct MM " Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 05/11] mm: debug: dump page " Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-06-30 23:35   ` David Rientjes
2015-06-30 23:35     ` David Rientjes
2015-07-01  8:53     ` Kirill A. Shutemov
2015-07-01  8:53       ` Kirill A. Shutemov
2015-07-01 19:21     ` Sasha Levin [this message]
2015-07-01 19:21       ` Sasha Levin
2015-07-01 21:25       ` David Rientjes
2015-07-01 21:25         ` David Rientjes
2015-07-01 21:34         ` Kirill A. Shutemov
2015-07-01 21:34           ` Kirill A. Shutemov
2015-07-01 22:33         ` Vlastimil Babka
2015-07-01 22:33           ` Vlastimil Babka
2015-07-01 22:50         ` Sasha Levin
2015-07-01 22:50           ` Sasha Levin
2015-07-08 23:58           ` David Rientjes
2015-07-08 23:58             ` David Rientjes
2015-08-06 15:08             ` Sasha Levin
2015-08-06 15:08               ` Sasha Levin
2015-05-14 17:10 ` [PATCH 06/11] mm: debug: clean unused code Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 07/11] mm: debug: VM_BUG() Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 08/11] mm: debug: kill VM_BUG_ON_PAGE Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 09/11] mm: debug: kill VM_BUG_ON_VMA Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 10/11] mm: debug: kill VM_BUG_ON_MM Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 17:10 ` [PATCH 11/11] mm: debug: use VM_BUG() to help with debug output Sasha Levin
2015-05-14 17:10   ` Sasha Levin
2015-05-14 20:24 ` [PATCH 00/11] mm: debug: formatting memory management structs Andrew Morton
2015-05-14 20:24   ` Andrew Morton
2015-05-14 20:26   ` Sasha Levin
2015-05-14 20:26     ` Sasha Levin
2015-06-26 21:34 ` Sasha Levin
2015-06-26 21:34   ` Sasha Levin

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=55943DC1.6010209@oracle.com \
    --to=sasha.levin@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=kirill@shutemov.name \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=rientjes@google.com \
    /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.