--
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |rth at gcc dot gnu dot org
|dot org |
Status|NEW
--- Additional Comments From pcarlini at suse dot de 2005-03-24 15:52
---
I'm recategorizing back to middle-end: really this should be properly fixed
in the middle-end.
--
What|Removed |Added
---
--- Additional Comments From pcarlini at suse dot de 2005-03-24 09:55
---
Argh! No, I was totally wrong: I tried fixing this problem some time ago and
did something wrong. We can fix it in the library. Sorry.
--
What|Removed |Added
---
--- Additional Comments From pcarlini at suse dot de 2005-03-24 09:33
---
Side note: here we are talking about the specializations for real/double/long
double, therefore no _M_real but __real__ _M_value and so on. In this case we
rely on the compiler to expand the operations using to bui
--- Additional Comments From pcarlini at suse dot de 2005-03-24 09:28
---
> It would be nice if the multiplication worked like this also for
> complex, even without -ffast-math. Or, is there something in the
> standard which would disallow this?
I don't think there is. But, AFAIK, thi
--- Additional Comments From fredrik dot huss at home dot se 2005-03-24
09:21 ---
Thanks for looking into this!
Yes, I was meaning when -ffast-math is NOT used, so maybe this is completely
unrelated. But I was thinking that even without -ffast-math, this should not
require a full comp
--
What|Removed |Added
CC||pcarlini at suse dot de
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20610
--- Additional Comments From pcarlini at suse dot de 2005-03-24 00:26
---
For concreteness, this is what I get (with 4.0.0 20050321) if I add
-ffast-math to your switches, seems not so bad, first blush:
_Z1fv:
.LFB1939:
pushl %ebp
.LCFI0:
movl%esp, %ebp
.LCFI1:
--- Additional Comments From pcarlini at suse dot de 2005-03-24 00:19
---
But, isn't 19953 about -ffast-math? Or you really want the same code *without*
that switch?!? We are talking about completely different issues, IMHO.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20610
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-23
21:11 ---
Because C99 is different from C++.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20610
10 matches
Mail list logo