Re: [patch] Fix vdso32 relocation for paddr

[AmÃrico Wang]

On Thu, Aug 27, 2009 at 6:31 AM, Jan
Kratochvil<jan.kratochvil@xxxxxxxxxx> wrote:
> Hi,
>
> `readelf -Wl file-dumped-from-/proc/self/maps-[vdso]' for x86_64 process prints:
> Program Headers:
> ÂType      Offset  VirtAddr      PhysAddr      FileSiz ÂMemSiz  Flg Align
> ÂLOAD Â Â Â Â Â 0x000000 0xffffffffff700000 0xffffffffff700000 0x000890 0x000890 R E 0x1000
> ÂDYNAMIC Â Â Â Â0x000410 0xffffffffff700410 0xffffffffff700410 0x0000f0 0x0000f0 R Â 0x8
> ÂNOTE Â Â Â Â Â 0x0002d0 0xffffffffff7002d0 0xffffffffff7002d0 0x00003c 0x00003c R Â 0x4
> ÂGNU_EH_FRAME Â 0x00030c 0xffffffffff70030c 0xffffffffff70030c 0x000034 0x000034 R Â 0x4
>
> while such dump for i386 process prints:
> Program Headers:
> ÂType      Offset  VirtAddr  PhysAddr  FileSiz MemSiz ÂFlg Align
> ÂLOAD Â Â Â Â Â 0x000000 0xffffe000 0x00000000 0x00434 0x00434 R E 0x1000
> ÂDYNAMIC Â Â Â Â0x000350 0xffffe350 0x00000350 0x00078 0x00078 R Â 0x4
> ÂNOTE Â Â Â Â Â 0x0001c4 0xffffe1c4 0x000001c4 0x00060 0x00060 R Â 0x4
> ÂGNU_EH_FRAME Â 0x000224 0xffffe224 0x00000224 0x00024 0x00024 R Â 0x4
>
> The mapped location can be arbirary (according to the rules of PIC - Position
> Independent Code) but currently the i386 vdso (vdso32) looks as an overlayed
> code as it has LMA != VMA (rule contained in GDB function section_is_overlay,
> as suggested by Daniel Jacobowitz).
>
> Fixed vdso32 prints:
> Program Headers:
> ÂType      Offset  VirtAddr  PhysAddr  FileSiz MemSiz ÂFlg Align
> ÂLOAD Â Â Â Â Â 0x000000 0xffffe000 0xffffe000 0x00434 0x00434 R E 0x1000
> ÂDYNAMIC Â Â Â Â0x000350 0xffffe350 0xffffe350 0x00078 0x00078 R Â 0x4
> ÂNOTE Â Â Â Â Â 0x0001c4 0xffffe1c4 0xffffe1c4 0x00060 0x00060 R Â 0x4
> ÂGNU_EH_FRAME Â 0x000224 0xffffe224 0xffffe224 0x00024 0x00024 R Â 0x4
>
> It required now a workaround in GDB patch so at least to fix it in Linux kernel:
> Â Â Â Âhttp://sourceware.org/ml/gdb-patches/2009-08/msg00437.html
>
>

Reviewed-by: WANG Cong <xiyou.wangcong@xxxxxxxxx>

This patch looks fine for me. But I think we should hear from Roland.
Roland?


> Thanks,
> Jan
>
>
> Signed-off-by: Jan Kratochvil <jan.kratochvil@xxxxxxxxxx>
>
> --- linux-2.6.30.x86_64/arch/x86/vdso/vdso32-setup.c-orig    2009-08-26 22:23:08.000000000 +0200
> +++ linux-2.6.30.x86_64/arch/x86/vdso/vdso32-setup.c  Â2009-08-26 22:38:10.000000000 +0200
> @@ -172,6 +172,7 @@ static __init void relocate_vdso(Elf32_E
> Â Â Â Âphdr = (void *)ehdr + ehdr->e_phoff;
> Â Â Â Âfor (i = 0; i < ehdr->e_phnum; i++) {
> Â Â Â Â Â Â Â Âphdr[i].p_vaddr += VDSO_ADDR_ADJUST;
> + Â Â Â Â Â Â Â phdr[i].p_paddr += VDSO_ADDR_ADJUST;
>
> Â Â Â Â Â Â Â Â/* relocate dynamic stuff */
> Â Â Â Â Â Â Â Âif (phdr[i].p_type == PT_DYNAMIC)
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at Âhttp://vger.kernel.org/majordomo-info.html
> Please read the FAQ at Âhttp://www.tux.org/lkml/
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/