Aaron Durbin (adurbin@google.com) just uploaded a new patch set to gerrit, which you can find at http://review.coreboot.org/2865
-gerrit
commit 504fb0ae5384f1b38a95b0bf4aae2cc393a36435 Author: Aaron Durbin adurbin@chromium.org Date: Wed Mar 20 13:49:27 2013 -0500
x86: don't clear bss in ramstage entry
The cbfs stage loading routine already zeros out the full memory region that a stage will be loaded. Therefore, it is unnecessary to to clear the bss again after once ramstage starts.
Change-Id: Icc7021329dbf59bef948a41606f56746f21b507f Signed-off-by: Aaron Durbin adurbin@chromium.org --- src/arch/x86/lib/c_start.S | 11 ----------- src/lib/cbfs.c | 1 + 2 files changed, 1 insertion(+), 11 deletions(-)
diff --git a/src/arch/x86/lib/c_start.S b/src/arch/x86/lib/c_start.S index 295283b..35bc26b 100644 --- a/src/arch/x86/lib/c_start.S +++ b/src/arch/x86/lib/c_start.S @@ -29,17 +29,6 @@ _start:
cld
- /** clear bss, which unlike the stack is zero by definition */ - leal _bss, %edi - movl $_ebss, %ecx - subl %edi, %ecx - jz .Lnobss - shrl $2, %ecx /* it is 32 bit aligned, right? */ - xorl %eax, %eax - rep - stosl -.Lnobss: - /** poison the stack. Code should not count on the * stack being full of zeros. This stack poisoning * recently uncovered a bug in the broadcast SIPI diff --git a/src/lib/cbfs.c b/src/lib/cbfs.c index abb95ab..c56f550 100644 --- a/src/lib/cbfs.c +++ b/src/lib/cbfs.c @@ -129,6 +129,7 @@ void * cbfs_load_stage(struct cbfs_media *media, const char *name) name, (uint32_t) stage->load, stage->memlen, stage->entry); + /* Stages rely the below clearing so that the bss is initialized. */ memset((void *) (uint32_t) stage->load, 0, stage->memlen);
if (cbfs_decompress(stage->compression,