All of lore.kernel.org
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: Peter Lieven <pl@kamp.de>, Juergen Gross <jgross@suse.com>,
	xen-devel <xen-devel@lists.xenproject.org>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Subject: Re: [Qemu-devel] Regression with commit 095497ffc66b7f031
Date: Fri, 15 Jul 2016 12:12:05 +0200	[thread overview]
Message-ID: <fa83b579-443e-0746-9263-fe1078abeb68@redhat.com> (raw)
In-Reply-To: <5788B5CC.8050305@kamp.de>



On 15/07/2016 12:07, Peter Lieven wrote:
> Am 15.07.2016 um 12:02 schrieb Paolo Bonzini:
>>
>> On 15/07/2016 10:47, Juergen Gross wrote:
>>> Nothing scaring and no real difference between working and not working
>>> variant.
>>>
>>> Meanwhile I've been digging a little bit deeper and found the reason:
>>> libxenstore is setting up a reader thread which is waiting for the
>>> watch to fire. With above commit the stack size of that thread (16kB)
>>> is too small. Setting it to 32kB made qemu work again.
>> This makes very little sense (not your fault)...  The commit doesn't
>> change stack usage at all, TLS should not be on the stack.
> 
> But we still allocate the VncPalette for every thread, right? Even
> if it has nothing todo with VNC.

Yes, I'm just trying to understand the root cause.  Which is that glibc
actually does carve out TLS space from the requested stack size.  That
means that a program that has a lot of TLS variables, or has big TLS
variables, will fail in weird ways.

So that's two reasons why your patch is okay. :)

Paolo

WARNING: multiple messages have this Message-ID (diff)
From: Paolo Bonzini <pbonzini@redhat.com>
To: Peter Lieven <pl@kamp.de>, Juergen Gross <jgross@suse.com>,
	xen-devel <xen-devel@lists.xenproject.org>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Subject: Re: Regression with commit 095497ffc66b7f031
Date: Fri, 15 Jul 2016 12:12:05 +0200	[thread overview]
Message-ID: <fa83b579-443e-0746-9263-fe1078abeb68@redhat.com> (raw)
In-Reply-To: <5788B5CC.8050305@kamp.de>



On 15/07/2016 12:07, Peter Lieven wrote:
> Am 15.07.2016 um 12:02 schrieb Paolo Bonzini:
>>
>> On 15/07/2016 10:47, Juergen Gross wrote:
>>> Nothing scaring and no real difference between working and not working
>>> variant.
>>>
>>> Meanwhile I've been digging a little bit deeper and found the reason:
>>> libxenstore is setting up a reader thread which is waiting for the
>>> watch to fire. With above commit the stack size of that thread (16kB)
>>> is too small. Setting it to 32kB made qemu work again.
>> This makes very little sense (not your fault)...  The commit doesn't
>> change stack usage at all, TLS should not be on the stack.
> 
> But we still allocate the VncPalette for every thread, right? Even
> if it has nothing todo with VNC.

Yes, I'm just trying to understand the root cause.  Which is that glibc
actually does carve out TLS space from the requested stack size.  That
means that a program that has a lot of TLS variables, or has big TLS
variables, will fail in weird ways.

So that's two reasons why your patch is okay. :)

Paolo

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

  reply	other threads:[~2016-07-15 10:12 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-07-15  6:32 [Qemu-devel] Regression with commit 095497ffc66b7f031 Juergen Gross
2016-07-15  7:39 ` Peter Lieven
2016-07-15  7:39 ` [Qemu-devel] " Peter Lieven
2016-07-15  8:47   ` Juergen Gross
2016-07-15  9:03     ` Peter Lieven
2016-07-15  9:23       ` Juergen Gross
2016-07-15  9:23       ` Juergen Gross
2016-07-15  9:03     ` Peter Lieven
2016-07-15 10:02     ` [Qemu-devel] " Paolo Bonzini
2016-07-15 10:07       ` Peter Lieven
2016-07-15 10:07       ` [Qemu-devel] " Peter Lieven
2016-07-15 10:12         ` Paolo Bonzini [this message]
2016-07-15 10:12           ` Paolo Bonzini
2016-07-15 10:13           ` Peter Lieven
2016-07-15 10:13           ` [Qemu-devel] " Peter Lieven
2016-07-15 10:12       ` Gerd Hoffmann
2016-07-15 10:12         ` Gerd Hoffmann
2016-07-15 10:35         ` Paolo Bonzini
2016-07-15 10:35         ` [Qemu-devel] " Paolo Bonzini
2016-07-15 10:41           ` Juergen Gross
2016-07-15 12:42             ` Paolo Bonzini
2016-07-15 13:21               ` Juergen Gross
2016-07-15 13:21               ` [Qemu-devel] [Xen-devel] " Juergen Gross
2016-07-15 12:42             ` Paolo Bonzini
2016-07-15 10:41           ` Juergen Gross
2016-07-15 10:02     ` Paolo Bonzini
2016-07-15  8:47   ` Juergen Gross

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=fa83b579-443e-0746-9263-fe1078abeb68@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=jgross@suse.com \
    --cc=kraxel@redhat.com \
    --cc=pl@kamp.de \
    --cc=qemu-devel@nongnu.org \
    --cc=xen-devel@lists.xenproject.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.