Add -fno-stack-protector to the GCC options.
Thanks, Ward.
Hmmm - buildrom should have already been taking care of this. In script/build.settings, we do the kernel trick for trying -fno-stack-protector and set STACKPROTECT if it doesn't make it through, and that gets passed through to CPU_OPT in linuxbios.inc. I guess maybe I need to overload a different variable?
Jordan
On 05/09/07 18:41 -0400, Ward Vandewege wrote:
Add -fno-stack-protector to the GCC options.
Thanks, Ward.
-- Ward Vandewege ward@fsf.org Free Software Foundation - Senior System Administrator
Modern distros seem to need this.
Signed-off-by: Ward Vandewege ward@gnu.org
Index: packages/linuxbios/patches/no-stack-protector.patch
--- packages/linuxbios/patches/no-stack-protector.patch (revision 0) +++ packages/linuxbios/patches/no-stack-protector.patch (revision 0) @@ -0,0 +1,13 @@ +Index: src/mainboard/gigabyte/m57sli/Options.lb +=================================================================== +--- LinuxBIOSv2.orig/src/mainboard/gigabyte/m57sli/Options.lb (revision 2742) ++++ LinuxBIOSv2/src/mainboard/gigabyte/m57sli/Options.lb (working copy) +@@ -297,7 +297,7 @@
- ##
- ## The default compiler
- ##
+-default CC="$(CROSS_COMPILE)gcc -m32" ++default CC="$(CROSS_COMPILE)gcc -m32 -fno-stack-protector"
- default HOSTCC="gcc"
- ##
-- linuxbios mailing list linuxbios@linuxbios.org http://www.linuxbios.org/mailman/listinfo/linuxbios
On Wed, Sep 05, 2007 at 08:50:00PM -0600, Jordan Crouse wrote:
Hmmm - buildrom should have already been taking care of this. In script/build.settings, we do the kernel trick for trying -fno-stack-protector and set STACKPROTECT if it doesn't make it through, and that gets passed through to CPU_OPT in linuxbios.inc. I guess maybe I need to overload a different variable?
Very strange, I just removed the -fno-stack-protector patch and it still works, the CPU_OPT code is being passed through properly. But I definitely ran into this a few weeks ago when I started working on buildrom. I'm not sure what happened here; maybe I was working against an old version of the code.
So; please disregard this patch.
I still have 3 final patches pending for this changeset; I'm trying to clarify the licensing situation of the linux-tiny patches (http://elinux.org/Linux_Tiny) before I post them to this list.
Thanks, Jordan! Ward.