Control: reopen -1 Hi Anton,
On Sat, Feb 20, 2021 at 12:59:17PM +0000, Anton Ivanov wrote: > On 20/02/2021 10:33, Debian Bug Tracking System wrote: > > This is an automatic notification regarding your Bug report > > which was filed against the src:linux package: > > > > #940821: linux-image-5.2.0-2-amd64: file cache corruption with nfs4 > > > > It has been closed by Bastian Blank <wa...@debian.org>. > > > > Their explanation is attached below along with your original report. > > If this explanation is unsatisfactory and you have not received a > > better one in a separate message then please contact Bastian Blank > > <wa...@debian.org> by > > replying to this email. > > > > > I missed the question. Probably hit the spam bucket for some reason. > > I am able to reproduce it with more recent versions as well. > > The most recent one I have around is 5.4.0-0.bpo.2-amd64 > > Still reproducible 100% - just tested it. > > It is trivial to reproduce if anyone actually bothers to do so. Just grab a > big enough tree where make runs truly in parallel - openwrt is best, but > even the Linux kernel does the job. > > Mount it via nfs4 from another server (it will work even locally, but takes > longer to reproduce - may take a whole afternoon) > > Run while make -j 12 clean && make -j 12 ; do true ; done > > Leave it to run. On 6 cores/12 threads it takes 2-3 builds of openwrt or ~ > 5-8 linux kernel builds to blow up. More cores - faster. Less cores slower. > > I sent it to the mailing list too, but nobody could be bothered to even ask > any questions. Let's reopen then but can you try to followup with upstream? I think that's the best place to handle the issue. 5.4.0-0.bpo.2-amd64 is in meanwhile as well quite behind, can you still reproduce it with the current stable series? Most ucrrent in backports is 5.10.13 based. Regards, Salvatore