[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: maphash
|
From: |
Andrea Corallo |
|
Subject: |
Re: maphash |
|
Date: |
Thu, 03 Oct 2024 07:18:36 -0400 |
|
User-agent: |
Gnus/5.13 (Gnus v5.13) |
Mattias Engdegård <mattias.engdegard@gmail.com> writes:
> 1 okt. 2024 kl. 20.36 skrev Andrea Corallo <acorallo@gnu.org>:
>
>>> Would you also take a look at the four DOHASH_SAFE uses in
>>> comp.c, and see if there is any possibility of mutation of the tables
>>> iterated over? Then they could be changed to DOHASH.
>>
>> Thanks I think the use of DOHASH_SAFE is correct there.
>
> Mind explaining why? Note that DOHASH does attempt to catch some table
> mutations when checking is enabled.
>
> DOHASH_SAFE is intended for Fmapcar, where there may be mutations and we
> mustn't crash because of them but we don't care about semantics in that case.
Sorry from your prev message I understood the opposit. With abb932290df
we now use DOHASH (not that I expect any measurable perf impact).
Thanks
Andrea