1 |
On Tue, Feb 26, 2013 at 8:33 AM, Richard Yao <ryao@g.o> wrote: |
2 |
> Unless a significant issue is found in jemalloc itself, I do not see any |
3 |
> reason to continue using glibc's ptmalloc over jemalloc. As far as I |
4 |
> know, FreeBSD, NetBSD, Facebook and others are using jemalloc, so I |
5 |
> expect that no significant issues will be found. jemalloc will still |
6 |
> needs testing before we can consider it. Doing good tests is hard, so I |
7 |
> would like to crowd source ideas on the appropriate way to test a malloc |
8 |
> implementation from the list. My expectation is that people on the list |
9 |
> will collectively think of better ideas than I could on my own. |
10 |
> |
11 |
> With that said, what do people think? |
12 |
> |
13 |
|
14 |
Just want to point out a problem I have personally encountered with |
15 |
another alternate malloc (tcmalloc). Essentially, when you combine |
16 |
libGL.so from nvidia-drivers with tcmalloc, the Google Chrome web |
17 |
browser fails to properly close its worker processes. |
18 |
|
19 |
https://bugs.gentoo.org/show_bug.cgi?id=413637 |
20 |
|
21 |
This thread on the Chromium developer list goes into more detail: |
22 |
|
23 |
https://groups.google.com/a/chromium.org/forum/#!msg/chromium-dev/uLf5l669dCk/MXXfJfIvDF0J |
24 |
|
25 |
The main concern I would have is that upstream developers (like |
26 |
nvidia) are unlikely to test their software with an alternate malloc |
27 |
implementation, so we could end up running into some very subtle bugs. |
28 |
Are the saved cycles really worth it? |