|
Bill Allombert on Sat, 18 Oct 2025 11:03:05 +0200
|
[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]
|
Re: PARI/GP 2.17.2 testsuite hangs nondeterministically (macOS)
|
- To: Matthias Koeppe <mkoeppe@ucdavis.edu>
- Subject: Re: PARI/GP 2.17.2 testsuite hangs nondeterministically (macOS)
- From: Bill Allombert <Bill.Allombert@math.u-bordeaux.fr>
- Date: Sat, 18 Oct 2025 11:02:58 +0200
- Cc: pari-users@pari.math.u-bordeaux.fr
- Delivery-date: Sat, 18 Oct 2025 11:03:05 +0200
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/simple; d=math.u-bordeaux.fr; s=2022; t=1760778183; bh=Fqa3QSTnJuFLuBIlHFvdbs02yPRwNAtyjLjRec/1k3Q=; h=Date:From:To:Cc:Subject:References:In-Reply-To:From; b=ZZPaC4OP5CgzRoVT/aovuoBgvojvCda2LBc/zWPidnk6WvBPINhKc6dDyqu9YPKtq oH44mENgirj3qoi2ONb+AyP/909O7J8BYapNmwpR+OOl6nxiRO9cpiyN3jlQ4BWZB9 SiOOissevd/0gAZzh9mXiKYsg6PB5OcgdldyXaw6k7eZN0UiLHPPSunAQWW5XvXYaw sFvxVrUvP/xa1g9t6TxiM/cXK/oj1LksIfmJQkZ1/6G0UOPZPOyw5dslnla42xj2qW +mY+lh732MlclrYnH64efvgmUPQwamtlFfX7x871IdhtLOV2/DzFOBZXMjDl+Ktwrq 0/xBObg3CDdPuaC5iCSoN98fbGedce7NIcHsrApevPuW1Sfmx3I2jburHljjC6GMP/ a7GpnUr6F7n/nYUz9LnWPJgTnD8uHy9SN/cODodjKgSii91k2bU7UNr0vmpWUiy8U5 or5BXAtMf3mFIwgP9/6wdwPAPpT+D9hJjPVJxAhrh9WFTM2rbwhk2oC+TYHCqESpRT Z9TWFTxZZ1gQBDnt62BThch82wqQ4WO5ILKsQLqSH0c8oX0oU/O6XvkjPBl330WaDF H+kJR4YFQr3wd10sA0eyQc0eeuSCZ3hC3i18d85JxlE/OoasC57DPrdoZZtWQBc9XK 1fjgXRbufx24hv1vOYmGiPBs=
- In-reply-to: <CAJ_wo5gerNb2e8uPo9TUseafELpEZz4wgW+9FkoyHdwO5cLgwQ@mail.gmail.com>
- Mail-followup-to: Matthias Koeppe <mkoeppe@ucdavis.edu>, pari-users@pari.math.u-bordeaux.fr
- References: <CAJ_wo5gerNb2e8uPo9TUseafELpEZz4wgW+9FkoyHdwO5cLgwQ@mail.gmail.com>
On Fri, Oct 17, 2025 at 03:18:35PM -0700, Matthias Koeppe wrote:
> I am encountering nondeterministic hangs of PARI/GP while running its
> testsuite on macOS. It seems to happen in particular in higher CPU
> load situations.
> The hang appears to be a pthread deadlock. In both of the tests that I
> have seen affected by it (bnf, ellrank), the backtrace shows
> "ZM_gauss", via "gen_inccrt_i". More detail:
((Are you sure it is ellrank ? It looks more like ellnf))
Clearly it hangs in mt_queue_reset while waiting for threads to terminate.
> https://github.com/passagemath/passagemath/issues/1684
This report does not include any detail on the system platform.
(what is the CPU, the OS version, the libc version, the compiler version etc.)
We run the full PARI test suite with --mt=pthread on aarch64 MacOS 12.0 on a
M1 every day and we never noticed any hangs. Maybe they happen without us
noticing thought. The latest buildlog is
<https://pari.math.u-bordeaux.fr/buildlogs/caa3445f-35dd-4ced-86be-e830cd259888>
PARI does not use OpenMP so setting OMP_NUM_THREADS should not have any effect.
Cheers,
Bill.