https://sourceware.org/bugzilla/show_bug.cgi?id=16808
--- Comment #5 from Que Quotion ---
Marking my own bug invalid. This was a misunderstanding.
Somehow I'd gotten away with feeding the linker flags like -Ofast before, but
had I read the documentation I might have realized this doesn't make se
https://sourceware.org/bugzilla/show_bug.cgi?id=16808
--- Comment #4 from Cary Coutant ---
> Is the same true for ld.gold? In which case -O, -Os, -Og, -Ofast are not
> implemented and -O1 = -O2 = -O3 = -O4 and so on?
At -O1 and above, gold will use a higher compression level with
--compress-debu
https://sourceware.org/bugzilla/show_bug.cgi?id=16808
Markus Trippelsdorf changed:
What|Removed |Added
Status|NEW |RESOLVED
CC|
https://sourceware.org/bugzilla/show_bug.cgi?id=16808
--- Comment #2 from Que Quotion ---
Changed this option in LDFLAGS from -Ofast to -O4 and compilation proceeds as
normal.
--
You are receiving this mail because:
You are on the CC list for the bug.
__
https://sourceware.org/bugzilla/show_bug.cgi?id=16808
--- Comment #1 from quequotion at mailinator dot com ---
The manpage for ld says:
-O level
If level is a numeric values greater than zero ld optimizes the output.
This might take significantly longer and therefore probably should only be
e
