[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[MIT-Scheme-devel] floating-point environment in threads
From: |
Taylor R Campbell |
Subject: |
[MIT-Scheme-devel] floating-point environment in threads |
Date: |
Tue, 28 May 2013 19:44:10 +0000 |
User-agent: |
IMAIL/1.21; Edwin/3.116; MIT-Scheme/9.1.1 |
Date: Mon, 27 May 2013 17:13:20 -0700
From: Matt Birkholz <address@hidden>
Create-thread also breaks this, ever since someone added the near-
useless, ABI-dependent floenvs and stuck one in each thread.
Attached is an unfinished patch to use lazy switching for the
floating-point environment so that threads that don't modify their
floating-point environments can store #F there instead. I intend to
finish debugging it and commit it this week.
Some of the people who use the floating-point environment consider
putting memory-unsafe C libraries into Scheme's address space to be
useless, but let's not throw stones in this glass house of a project,
OK? It's already limping along slowly enough.
lazy-floenv.patch
Description: Text document
- [MIT-Scheme-devel] C callbacks into Scheme [was Re: sqlite3], (continued)
- [MIT-Scheme-devel] sqlite3, Matt Birkholz, 2013/05/18
- Re: [MIT-Scheme-devel] sqlite3, Taylor R Campbell, 2013/05/18
- Re: [MIT-Scheme-devel] sqlite3, Matt Birkholz, 2013/05/19
- Re: [MIT-Scheme-devel] sqlite3, Taylor R Campbell, 2013/05/19
- Re: [MIT-Scheme-devel] sqlite3, Matt Birkholz, 2013/05/19
- Re: [MIT-Scheme-devel] sqlite3, Taylor R Campbell, 2013/05/19
- Re: [MIT-Scheme-devel] sqlite3, Matt Birkholz, 2013/05/27
- [MIT-Scheme-devel] floating-point environment in threads,
Taylor R Campbell <=