Linux Trace Development
 help / color / mirror / Atom feed
From: Frank Heimes <frank.heimes@canonical.com>
To: linux-trace-devel@vger.kernel.org
Cc: Steven Rostedt <rostedt@goodmis.org>,
	Benjamin Drung <benjamin.drung@canonical.com>,
	Jamie Dolan <jldolan@us.ibm.com>
Subject: [PATCH 0/3] libtracefs utest: Fix trace cpu pipe and test_mem failures
Date: Wed, 18 Nov 2026 13:30:50 +0000	[thread overview]
Message-ID: <cover.1795008446.git.frank.heimes@canonical.com> (raw)

These fix unit test failures reported here:

  https://lore.kernel.org/linux-trace-devel/39efe87bec205b7f809b8aa49db635b6465ee530.camel@canonical.com/

 - Patch 1: on architectures with 64K pages (ppc64el), the large step of
   "trace cpu pipe" writes more than 13 GB into /tmp with the memory
   mapped buffer, which fails with ENOSPC on a tmpfs /tmp.

 - Patch 2: "make test_mem" hangs in "trace pid events filter", as the
   spinning threads starve the main thread under valgrind.

 - Patch 3: "trace cpu pipe" fails under valgrind when the memory mapped
   reader can not keep up and the kernel overwrites events.

The patches are independent of each other. I can also send them
separately, if that's preferred.

Related fixes for the same report:

 - kbuffer: Fix out-of-bounds write in kbuffer_read_buffer()
   https://lore.kernel.org/linux-trace-devel/cover.1794988946.git.frank.heimes@canonical.com/T/#u
 - libtracefs: Do not miss sub-buffers when reading a memory mapped buffer
   https://lore.kernel.org/linux-trace-devel/20261118102644.3392133-1-frank.heimes@canonical.com/T/#u
 - libtracefs utest: Fix buffer size in follow events clear test
   https://lore.kernel.org/linux-trace-devel/20261116235650.21314-1-jldolan@us.ibm.com/

Tested on ppc64el (POWER9, 64K pages, kernel 7.3) on top of master,
together with the related fixes above: "trace cpu pipe" passes in "make
test" and "make test_mem", and "make test_mem" no longer hangs.

Frank Heimes (3):
  libtracefs utest: Scale trace cpu pipe test to the sub-buffer size
  libtracefs utest: Run valgrind memory test with fair scheduling
  libtracefs utest: Account for overruns in the trace cpu pipe test

 Makefile              |  1 +
 utest/tracefs-utest.c | 42 ++++++++++++++++++++++++++++++++++++++++--
 2 files changed, 41 insertions(+), 2 deletions(-)

-- 
2.53.0


             reply	other threads:[~2026-11-18 13:31 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-18 13:30 Frank Heimes [this message]
2026-11-18 13:30 ` [PATCH 1/3] libtracefs utest: Scale trace cpu pipe test to the sub-buffer size Frank Heimes
2026-11-18 13:30 ` [PATCH 2/3] libtracefs utest: Run valgrind memory test with fair scheduling Frank Heimes
2026-11-18 13:30 ` [PATCH 3/3] libtracefs utest: Account for overruns in the trace cpu pipe test Frank Heimes

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=cover.1795008446.git.frank.heimes@canonical.com \
    --to=frank.heimes@canonical.com \
    --cc=benjamin.drung@canonical.com \
    --cc=jldolan@us.ibm.com \
    --cc=linux-trace-devel@vger.kernel.org \
    --cc=rostedt@goodmis.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