On 10 Feb 2025, at 16:21, Matthew Wilcox (Oracle) wrote: > syzbot has reported a number of oopses caused by reading > /proc/kpageflags racing with a folio being split / freed / allocated > and thus a page which starts out as a hed page becomes a tail page s/hed/head/ > during the read, which our assertions catch as an error. > > To solve this problem, snapshot the page like dump_page() does. > > Link: the syzbot reports > Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx> > --- > fs/proc/page.c | 24 ++++++++++++++---------- > 1 file changed, 14 insertions(+), 10 deletions(-) LGTM. Reviewed-by: Zi Yan <ziy@xxxxxxxxxx> Best Regards, Yan, Zi
