[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Weak tables harmful to GC?
From: |
Ludovic Courtès |
Subject: |
Re: Weak tables harmful to GC? |
Date: |
Sat, 21 Oct 2017 18:52:32 -0700 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/25.3 (gnu/linux) |
Hello!
address@hidden (Ludovic Courtès) skribis:
> All in all, given that these issues are very likely causes of the
> execution time and memory consumption issues that plague the compiler
> (where we have huge symbol and source property tables), I’m in favor of
> switching back to the 2.0 implementation of weak hash tables. That can
> be done in an API-compatible way, I think.
The patch below reverts to the 2.0 bucket-and-weak-lists weak hash
tables.
On my benchmark, which calls ‘emit-bytecode’ on a huge piece of code
(corresponding to (gnu packages python)), it runs ~18% faster than
current 2.2; memory usage is reduced by only ~3% at the end.
It appears to work well except under stress: a tight loop as in
<https://bugs.gnu.org/28590> easily triggers the “weak hash table
corruption” message or an assertion failure.
To be continued…
Ludo’.
0001-Remove-weak-tables-and-revert-to-weak-hash-tables.patch
Description: Text Data
- Re: Weak tables harmful to GC?, Christopher Allan Webber, 2017/10/09
- Re: Weak tables harmful to GC?,
Ludovic Courtès <=
- Re: Weak tables harmful to GC?, Christopher Allan Webber, 2017/10/22
- Re: Weak tables harmful to GC?, Ludovic Courtès, 2017/10/22
- Re: Weak tables harmful to GC?, Christopher Allan Webber, 2017/10/24
- Re: Weak tables harmful to GC?, Ludovic Courtès, 2017/10/24
- Re: Weak tables harmful to GC?, Christopher Allan Webber, 2017/10/24
- Re: Weak tables harmful to GC?, Ludovic Courtès, 2017/10/25
- Re: Weak tables harmful to GC?, Ricardo Wurmus, 2017/10/25
- Re: Weak tables harmful to GC?, Ricardo Wurmus, 2017/10/25
- Re: Weak tables harmful to GC?, Ludovic Courtès, 2017/10/26
- Re: Weak tables harmful to GC?, Ricardo Wurmus, 2017/10/26