On 02/06/2026 11:26, Shung-Hsi Yu wrote:
On Tue, Apr 21, 2026 at 07:59:08AM +0200, Jerome Marchand wrote:
On 4/20/26 6:52 PM, Mykyta Yatsenko wrote:
On 4/20/26 2:46 PM, Jerome Marchand wrote:
The file_reader/on_open_expect_fault fails consistently on my system.
It expects a page fault on first dynptr read of some range the exe
file of the current process because it has paged out that page range
earlier. However a lot can happen to that range (which depending on
the actual memory layout could contain text section, data section,
sections )related to dynamic linking...) between the moment it was
paged out and the moment the bpf program expected to hit a pagefault
actually run.
A bit of instrumentation with mincore() shows that pages from that
range were accessed several times before the program is run. In
particular the call of file_reader__load() seems to fault all the
range in.
Move the call to madvise(MADV_PAGEOUT) to just before attaching the
program to minimize the risk of having those page pulled back in from
under our feet.
Signed-off-by: Jerome Marchand <[email protected]>
---
Thank you for the patch, the change looks good. Does it fail
consistently on 4K page size?
It did when I ran the test manually. On our automated testing system, it
failed intermittently.
On my fork of BPF CI the test was failing[1] consistently, even when
this commit is applied. However I wasn't able to reproduce locally on my
laptop when I ran manually.
Any hints regarding where to look and how to reproduce?
To investigate the failure I witnessed I instrumented the test with some
calls to mincore(). You can do that to see whether madvise(PAGEOUT)
didn't successfully paged out those pages, or if they were paged back in
by the MM subsystem. I'm afraid that this test isn't very robust: the MM
subsystem is the one that ultimately decides which pages are in or out
and it might not work in the way that this test assumes.
Thanks,
Jerome
Thanks,
Shung-Hsi
1: https://github.com/shunghsiyu/libbpf/actions/runs/26807198259/job/79027625188
...
