Author inada.naoki
Recipients benjamin.peterson, fweimer, gregory.p.smith, inada.naoki, nascheme, pitrou, skrah, tgrigg, twouters, vstinner
Date 2019-04-16.10:14:54
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1555409694.37.0.933110224573.issue27987@roundup.psfhosted.org>
In-reply-to
Content
> In 3.8, the union used to ensure alignment on a C double is gone.

Note that two uintptr_t is aligned 16bytes on 64bit platforms and 8bytes on 32bit platforms.

Python 3.7 is worse than 3.8.
It used "double dummy" to align by 8 bytes, not 16 bytes.
We should use "long double" to align by 16 bytes.
https://software.intel.com/sites/default/files/article/402129/mpx-linux64-abi.pdf

But it means +8 bytes for all tuples.  If we backport PR-12850 to 3.7, +8 bytes for 1/2 tuples, and +16 bytes for remaining tuples.

Any ideas about reduce impact for Python 3.7?
For example, can we add 8byte dummy to PyGC_Head, and tuple use the dummy for hash?  Maybe, it breaks ABI....  Not a chance...

I wonder if we can add -fmax-type-align=8 for extension types...
History
Date User Action Args
2019-04-16 10:14:54inada.naokisetrecipients: + inada.naoki, twouters, nascheme, gregory.p.smith, pitrou, vstinner, benjamin.peterson, skrah, fweimer, tgrigg
2019-04-16 10:14:54inada.naokisetmessageid: <1555409694.37.0.933110224573.issue27987@roundup.psfhosted.org>
2019-04-16 10:14:54inada.naokilinkissue27987 messages
2019-04-16 10:14:54inada.naokicreate