| 1 |
On Fri, 10 Jan 2014 15:08:02 -0500 |
| 2 |
"Anthony G. Basile" <basile@××××××××××××××.edu> wrote: |
| 3 |
|
| 4 |
> On 01/10/2014 10:50 AM, Ryan Hill wrote: |
| 5 |
> > Having slept on it I'm starting to agree. My first argument was that on |
| 6 |
> > hardened ssp is -fstack-protector-all, which is much more expensive, and it |
| 7 |
> > adds -fstack-check and -z,now to the linker by default as well. The pie |
| 8 |
> > half |
| 9 |
> |
| 10 |
> I'm pretty sure we're not adding -fstack-check unless something has |
| 11 |
> changed. Where are you seeing that? |
| 12 |
> |
| 13 |
> The reason I'm concerned is because of situations like bug #471756. |
| 14 |
> stack-check incumbers a register which in some situations (like the asm |
| 15 |
> in ffmpeg) can get you into trouble with not enough GENERAL_REGS. |
| 16 |
|
| 17 |
40_all_gcc48_config_esp.patch: |
| 18 |
|
| 19 |
54 + #if defined ( EFAULT_SSP ) || defined ( EFAULT_PIE_SSP ) |
| 20 |
55 + #define ESP_OPTIONS_SSP_SPEC \ |
| 21 |
56 + "%{nostdlib|nodefaultlibs|ffreestanding|fno-stack-protector| \ |
| 22 |
57 + fstack-protector|fstack-protector-all:;:-fstack-protector-all} \ |
| 23 |
58 + %{fstack-check|fstack-check=*:;: -fstack-check}" |
| 24 |
|
| 25 |
> > It might be useful to have these flags so we can mask them on archs that |
| 26 |
> > don't support ssp/pie. But that's always been true and it looks like sh is |
| 27 |
> > the only place we've bothered for some reason. |
| 28 |
> |
| 29 |
> Yes please. I had this issue on mips where gcc didn't support ssp for |
| 30 |
> early versions of gcc 4.x. |
| 31 |
|
| 32 |
There's a list of arches in every gcc ebuild that support PIE and SSP for |
| 33 |
both glibc and uclibc. AFAICT you would just have to remove mips from that |
| 34 |
list. |
| 35 |
|
| 36 |
|
| 37 |
-- |
| 38 |
Ryan Hill psn: dirtyepic_sk |
| 39 |
gcc-porting/toolchain/wxwidgets @ gentoo.org |
| 40 |
|
| 41 |
47C3 6D62 4864 0E49 8E9E 7F92 ED38 BD49 957A 8463 |
Archives