Messages in this thread Patch in this message |  | | Date | Thu, 15 Nov 2001 20:13:40 +1100 | From | Anton Blanchard <> | Subject | [PATCH] reiserfs gcc alias bug only on ppc32 |
| |
Hi,
The workaround in reiserfs for a gcc bug in some ppc32 compilers does not affect the ppc64 compiler. A quick test to see stack usage for both options shows:
gcc -O1 reiserfs_delete_solid_item: 976 reiserfs_rename: 960 reiserfs_cut_from_item: 880
gcc -O2 reiserfs_rename: 1008 reiserfs_delete_solid_item: 976 reiserfs_cut_from_item: 896
So there is no reason to compile reiserfs using -O1 on ppc64.
Anton
--- 2.4.15-pre4/fs/reiserfs/Makefile Thu Nov 15 20:06:13 2001 +++ 2.4.15-pre4_work/fs/reiserfs/Makefile Thu Nov 15 20:05:26 2001 @@ -13,13 +13,13 @@ obj-m := $(O_TARGET) -# gcc -O2 (the kernel default) is overaggressive on ppc when many inline +# gcc -O2 (the kernel default) is overaggressive on ppc32 when many inline # functions are used. This causes the compiler to advance the stack # pointer out of the available stack space, corrupting kernel space, -# and causing a panic. Since this behavior only affects ppc, this ifeq +# and causing a panic. Since this behavior only affects ppc32, this ifeq # will work around it. If any other architecture displays this behavior, # add it here. -ifeq ($(shell uname -m),ppc) +ifeq ($(CONFIG_PPC32),y) EXTRA_CFLAGS := -O1 endif - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
|  |