Dan Kegel wrote:
If it turns out glibc's new __thread variable support really can do what you need on all platforms, do you agree that it might be better to use that?
Insert the word "new" between "all" and "platforms", and maybe you'll have more of an appreciation for my point of view :-)
Which platforms are those? I didn't see this in your previous email.
How about basically every IA32 Linux installation out there (or, at the very least, those that are currently supported by our drivers)? How many people are actually using systems that have full support for __thread? As I understand it, RedHat 8.1 will be the first large-scale release that does.
-- Gareth
Gareth Hughes wrote:
Dan Kegel wrote:
If it turns out glibc's new __thread variable support really can do what you need on all platforms, do you agree that it might be better to use that?
Insert the word "new" between "all" and "platforms", and maybe you'll have more of an appreciation for my point of view :-)
OK, "all new platforms". Sounds like a good argument for tagging along with glibc's __thread variable support, if you ask me.
- We need fast TLS access on platforms where __thread variables are not
supported.
Which platforms are those? I didn't see this in your previous email.
How about basically every IA32 Linux installation out there (or, at the very least, those that are currently supported by our drivers)? How many people are actually using systems that have full support for __thread? As I understand it, RedHat 8.1 will be the first large-scale release that does.
Yes, and all other Linux distros will follow not too long thereafter.
- Dan
-
Dan Kegel -
Gareth Hughes