qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Ahmed Karaman <1880722@bugs.launchpad.net>
To: qemu-devel@nongnu.org
Subject: [Bug 1880722] Re: Problems related to checking page crossing in use_goto_tb()
Date: Tue, 26 May 2020 17:00:57 -0000	[thread overview]
Message-ID: <159051245801.8264.18308114750913433257.launchpad@gac.canonical.com> (raw)
In-Reply-To: 159050902512.8264.16089666576245848412.malonedeb@gac.canonical.com

** Summary changed:

- Changing executable page permissions with mmap causes user-mode failures
+ Problems related to checking page crossing in use_goto_tb()

** Description changed:

- The discussion that led to this bug discovery can be found in this 
+ The discussion that led to this bug discovery can be found in this
  mailing list thread:
  https://lists.nongnu.org/archive/html/qemu-devel/2020-05/msg05426.html
  
- A workaround for this problem would be to check for page crossings for 
- both the user and system modes in the use_goto_tb() function across 
+ A workaround for this problem would be to check for page crossings for
+ both the user and system modes in the use_goto_tb() function across
  targets. Some targets like "hppa" already implement this fix but others
  don't.
  
- To solve the root cause of this problem, the linux-user/mmap.c should 
- be fixed to do all the invalidations required. By doing so, up to 6.93% 
- performance improvements will be achieved.
+ To solve the root cause of this problem, the linux-user/mmap.c should
+ be fixed to do all the invalidations required. By doing so, better
+ performance results could be achieved, compared to the case of the
+ workaround described above.

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1880722

Title:
  Problems related to checking page crossing in use_goto_tb()

Status in QEMU:
  New

Bug description:
  The discussion that led to this bug discovery can be found in this
  mailing list thread:
  https://lists.nongnu.org/archive/html/qemu-devel/2020-05/msg05426.html

  A workaround for this problem would be to check for page crossings for
  both the user and system modes in the use_goto_tb() function across
  targets. Some targets like "hppa" already implement this fix but others
  don't.

  To solve the root cause of this problem, the linux-user/mmap.c should
  be fixed to do all the invalidations required. By doing so, better
  performance results could be achieved, compared to the case of the
  workaround described above.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1880722/+subscriptions


  parent reply	other threads:[~2020-05-26 17:11 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-26 16:03 [Bug 1880722] [NEW] Changing executable page permissions with mmap causes user-mode failures Ahmed Karaman
2020-05-26 16:30 ` [Bug 1880722] " Ahmed Karaman
2020-05-26 17:00 ` Ahmed Karaman [this message]
2021-05-06 17:15 ` [Bug 1880722] Re: Problems related to checking page crossing in use_goto_tb() Thomas Huth
2021-07-06  4:17 ` Launchpad Bug Tracker

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=159051245801.8264.18308114750913433257.launchpad@gac.canonical.com \
    --to=1880722@bugs.launchpad.net \
    --cc=qemu-devel@nongnu.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).