Date: Mon, 16 Feb 2004 17:52:04 -0800 (PST) From: Matthew Dillon <dillon@apollo.backplane.com> To: kientzle@acm.org Cc: freebsd-hackers@freebsd.org Subject: Re: FreeBSD 5.2 v/s FreeBSD 4.9 MFLOPS performance (gcc3.3.3 v/sgcc2.9.5) Message-ID: <200402170152.i1H1q44u088726@apollo.backplane.com> References: <BAY12-F357RapPBVToy00031029@hotmail.com> <200402162112.i1GLCFMV087316@apollo.backplane.com> <4031678D.2060704@acm.org>
next in thread | previous in thread | raw e-mail | index | archive | help
:I've not looked at 3.3, but I seem to recall that GCC 3.2
:did not actually align the stack within each function, but
:preserved the alignment. (That is, each function assumed the stack
:had a certain alignment on entry and ensured that alignment
:was preserved for any subsequent function calls.)
Easy to test... ah, ok. 3.3 aligns the stack in main().
main:
pushl %ebp
movl %esp, %ebp
subl $8, %esp
andl $-16, %esp <<<<< ailgns stack here andl 0xfffffff0,%esp
...
And the preserves the alignment in other procedures... 8 + ebp + retaddr
is 16 bytes:
charlie:
pushl %ebp
movl %esp, %ebp
subl $8, %esp /* I declared 'volatile int x' as a stack var */
call fubar
call fubar
call fubar
leave
ret
:If I'm remembering this correctly, then aligning
:the stack in crt1.o would be pretty much essential.
:
:Tim Kientzle
For gcc 2.95, yes.
-Matt
Matthew Dillon
<dillon@backplane.com>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200402170152.i1H1q44u088726>