On Fri, Sep 28 2018, Ævar Arnfjörð Bjarmason wrote:
> On Thu, Sep 27 2018, Ævar Arnfjörð Bjarmason wrote: > >> On Thu, Sep 27 2018, SZEDER Gábor wrote: >> >>> On Thu, Sep 27, 2018 at 03:53:24PM +0200, Ævar Arnfjörð Bjarmason wrote: >>>> >>>> On Thu, Sep 27 2018, SZEDER Gábor wrote: >>>> >>>> > This is the second attempt to fix the racy split index problem, which >>>> > causes occasional failures in several random test scripts when run >>>> > with 'GIT_TEST_SPLIT_INDEX=yes'. The important details are in patches >>>> > 1 and 5 (corresponding to v1's 3 and 5). >>>> >>>> Thanks. I'm running the same sorts of tests I noted in >>>> https://public-inbox.org/git/87va7ireuu....@evledraar.gmail.com/ on >>>> this. The fix Jeff had that you noted in >>>> https://public-inbox.org/git/20180906151439.GA8016@localhost/ is now in >>>> "master". >>>> >>>> I take it your >>>> https://github.com/szeder/git/commits/racy-split-index-fix is the same >>>> as this submission? >>> >>> Yes. >>> >>>> Anyway, I'm testing that cherry-picked on top of the >>>> latest master. >>>> >>>> Unfortunate that we couldn't get the isolated test you made in >>>> https://public-inbox.org/git/20180907034942.GA10370@localhost/ >>> >>> Nah, that's not an isolated test case, that's only a somewhat >>> narrowed-down, but rather reliable reproduction recipe while I still >>> had no idea what was going on :) >>> >>> The _real_ isolated test is the last test in t1701, that's what it >>> eventually boiled down to. >> >> Thanks. I had ~400 runs of the tests I ran before and they were all >> OK. Now trying also with t1701 (which I hadn't noticed was a new >> test...). > > Ran that overnight with the same conditions as before. 2683 OK runs and > 0 failures (and counting). So it seems like the combination of the two > fixed the split index bugs. I forgot I ad this running, and got up to 45482 OKs and 0 FAILs before finally Ctrl+C-ing it now :) >>>> but I >>>> don't see how it could be added without some very liberal >>>> getenv("GIT_TEST_blahblah"), so it's probably best to not add it, >>>> particularly with the C rewrite of git-stash in-flight. >>>> >>>> I'll report back when I have enough test data to say how these patches >>>> affect the intermittent test failures under GIT_TEST_SPLIT_INDEX=yes.
