diff options
author | Ulrich Weigand <Ulrich.Weigand@de.ibm.com> | 2013-12-04 06:59:37 -0600 |
---|---|---|
committer | Adhemerval Zanella <azanella@linux.vnet.ibm.com> | 2013-12-04 07:41:39 -0600 |
commit | 61cd8fe4017c251617dd300818917e61a12ab48e (patch) | |
tree | dfac7e867271b7a5889db22693883cd54b41e9ac /sysdeps/powerpc/powerpc64/dl-trampoline.S | |
parent | 8b8a692cfd7d80f1ee7c8b9ab356a259367dd187 (diff) | |
download | glibc-61cd8fe4017c251617dd300818917e61a12ab48e.tar.gz glibc-61cd8fe4017c251617dd300818917e61a12ab48e.tar.xz glibc-61cd8fe4017c251617dd300818917e61a12ab48e.zip |
PowerPC64 ELFv2 ABI 5/6: LD_AUDIT interface changes
The ELFv2 ABI changes the calling convention by passing and returning structures in registers in more cases than the old ABI: http://gcc.gnu.org/ml/gcc-patches/2013-11/msg01145.html http://gcc.gnu.org/ml/gcc-patches/2013-11/msg01147.html For the most part, this does not affect glibc, since glibc assembler files do not use structure parameters / return values. However, one place is affected: the LD_AUDIT interface provides a structure to the audit routine that contains all registers holding function argument and return values for the intercepted PLT call. Since the new ABI now sometimes uses registers to return values that were never used for this purpose in the old ABI, this structure has to be extended. To force audit routines to be modified for the new ABI if necessary, the patch defines v2 variants of the la_ppc64 types and routines. In addition, the patch contains two unrelated changes to the PLT trampoline routines: it fixes a bug where FPR return values were stored in the wrong place, and it removes the unnecessary save/restore of CR.
Diffstat (limited to 'sysdeps/powerpc/powerpc64/dl-trampoline.S')
-rw-r--r-- | sysdeps/powerpc/powerpc64/dl-trampoline.S | 131 |
1 files changed, 89 insertions, 42 deletions
diff --git a/sysdeps/powerpc/powerpc64/dl-trampoline.S b/sysdeps/powerpc/powerpc64/dl-trampoline.S index 18c8a3aa01..69ce523e90 100644 --- a/sysdeps/powerpc/powerpc64/dl-trampoline.S +++ b/sysdeps/powerpc/powerpc64/dl-trampoline.S @@ -50,11 +50,8 @@ EALIGN(_dl_runtime_resolve, 4, 0) /* Store the LR in the LR Save area. */ std r0,FRAME_SIZE+FRAME_LR_SAVE(r1) cfi_offset (lr, FRAME_LR_SAVE) - mfcr r0 std r9,INT_PARMS+48(r1) std r10,INT_PARMS+56(r1) -/* I'm almost certain we don't have to save cr... be safe. */ - std r0,FRAME_SIZE+FRAME_CR_SAVE(r1) bl JUMPTARGET(_dl_fixup) #ifndef SHARED nop @@ -66,11 +63,9 @@ EALIGN(_dl_runtime_resolve, 4, 0) ld r8,INT_PARMS+40(r1) ld r7,INT_PARMS+32(r1) mtlr r0 - ld r0,FRAME_SIZE+FRAME_CR_SAVE(r1) ld r6,INT_PARMS+24(r1) ld r5,INT_PARMS+16(r1) ld r4,INT_PARMS+8(r1) - mtcrf 0xFF,r0 /* Prepare for calling the function returned by fixup. */ PPC64_LOAD_FUNCPTR r3 ld r3,INT_PARMS+0(r1) @@ -85,18 +80,30 @@ END(_dl_runtime_resolve) #undef FRAME_SIZE #undef INT_PARMS - /* Stack layout: - (Note: some of these are not required for the ELFv2 ABI.) - +592 previous backchain - +584 spill_r31 - +576 spill_r30 - +560 v1 - +552 fp4 - +544 fp3 - +536 fp2 - +528 fp1 - +520 r4 - +512 r3 + /* Stack layout: ELFv2 ABI. + +752 previous backchain + +744 spill_r31 + +736 spill_r30 + +720 v8 + +704 v7 + +688 v6 + +672 v5 + +656 v4 + +640 v3 + +624 v2 + +608 v1 + +600 fp10 + ELFv1 ABI +592 fp9 + +592 previous backchain +584 fp8 + +584 spill_r31 +576 fp7 + +576 spill_r30 +568 fp6 + +560 v1 +560 fp5 + +552 fp4 +552 fp4 + +544 fp3 +544 fp3 + +536 fp2 +536 fp2 + +528 fp1 +528 fp1 + +520 r4 +520 r4 + +512 r3 +512 r3 return values +504 free +496 stackframe @@ -157,10 +164,15 @@ END(_dl_runtime_resolve) +8 CR save area r1+0 stack back chain */ -#define FRAME_SIZE 592 +#if _CALL_ELF == 2 +# define FRAME_SIZE 752 +# define VR_RTN 608 +#else +# define FRAME_SIZE 592 +# define VR_RTN 560 +#endif #define INT_RTN 512 #define FPR_RTN 528 -#define VR_RTN 560 #define STACK_FRAME 496 #define CALLING_LR 488 #define CALLING_SP 480 @@ -205,18 +217,14 @@ EALIGN(_dl_profile_resolve, 4, 0) mflr r5 std r7,INT_PARMS+32(r1) std r8,INT_PARMS+40(r1) -/* Store the LR in the LR Save area of the previous frame. */ -/* XXX Do we have to do this? */ +/* Store the LR in the LR Save area. */ la r8,FRAME_SIZE(r1) std r5,FRAME_SIZE+FRAME_LR_SAVE(r1) cfi_offset (lr, FRAME_LR_SAVE) std r5,CALLING_LR(r1) - mfcr r0 std r9,INT_PARMS+48(r1) std r10,INT_PARMS+56(r1) std r8,CALLING_SP(r1) -/* I'm almost certain we don't have to save cr... be safe. */ - std r0,FRAME_SIZE+FRAME_CR_SAVE(r1) ld r12,.LC__dl_hwcap@toc(r2) #ifdef SHARED /* Load _rtld_local_ro._dl_hwcap. */ @@ -319,11 +327,9 @@ L(restoreFXR): ld r8,INT_PARMS+40(r1) ld r7,INT_PARMS+32(r1) mtlr r0 - ld r0,FRAME_SIZE+FRAME_CR_SAVE(r1) ld r6,INT_PARMS+24(r1) ld r5,INT_PARMS+16(r1) ld r4,INT_PARMS+8(r1) - mtcrf 0xFF,r0 /* Prepare for calling the function returned by fixup. */ PPC64_LOAD_FUNCPTR r3 ld r3,INT_PARMS+0(r1) @@ -346,10 +352,11 @@ L(restoreFXR): lfd fp12,FPR_PARMS+88(r1) lfd fp13,FPR_PARMS+96(r1) /* Unwind the stack frame, and jump. */ - ld r31,584(r1) - ld r30,576(r1) + ld r31,FRAME_SIZE-8(r1) + ld r30,FRAME_SIZE-16(r1) addi r1,r1,FRAME_SIZE bctr + L(do_pltexit): la r10,(VR_PARMS+0)(r1) la r9,(VR_PARMS+16)(r1) @@ -383,11 +390,9 @@ L(restoreFXR2): ld r8,INT_PARMS+40(r1) ld r7,INT_PARMS+32(r1) mtlr r0 - ld r0,FRAME_SIZE+FRAME_CR_SAVE(r1) ld r6,INT_PARMS+24(r1) ld r5,INT_PARMS+16(r1) ld r4,INT_PARMS+8(r1) - mtcrf 0xFF,r0 /* Prepare for calling the function returned by fixup. */ std r2,FRAME_TOC_SAVE(r1) PPC64_LOAD_FUNCPTR r3 @@ -413,16 +418,37 @@ L(restoreFXR2): /* But return here and store the return values. */ std r3,INT_RTN(r1) std r4,INT_RTN+8(r1) - stfd fp1,FPR_PARMS+0(r1) - stfd fp2,FPR_PARMS+8(r1) + stfd fp1,FPR_RTN+0(r1) + stfd fp2,FPR_RTN+8(r1) cmpdi cr0,r12,0 la r10,VR_RTN(r1) - stfd fp3,FPR_PARMS+16(r1) - stfd fp4,FPR_PARMS+24(r1) + stfd fp3,FPR_RTN+16(r1) + stfd fp4,FPR_RTN+24(r1) +#if _CALL_ELF == 2 + la r12,VR_RTN+16(r1) + stfd fp5,FPR_RTN+32(r1) + stfd fp6,FPR_RTN+40(r1) + li r5,32 + li r6,64 + stfd fp7,FPR_RTN+48(r1) + stfd fp8,FPR_RTN+56(r1) + stfd fp9,FPR_RTN+64(r1) + stfd fp10,FPR_RTN+72(r1) +#endif mr r3,r31 mr r4,r30 beq L(callpltexit) stvx v2,0,r10 +#if _CALL_ELF == 2 + stvx v3,0,r12 + stvx v4,r5,r10 + stvx v5,r5,r12 + addi r5,r5,64 + stvx v6,r6,r10 + stvx v7,r6,r12 + stvx v8,r5,r10 + stvx v9,r5,r12 +#endif L(callpltexit): addi r5,r1,INT_PARMS addi r6,r1,INT_RTN @@ -434,18 +460,39 @@ L(callpltexit): lwz r12,VR_VRSAVE(r1) ld r3,INT_RTN(r1) ld r4,INT_RTN+8(r1) - lfd fp1,FPR_PARMS+0(r1) - lfd fp2,FPR_PARMS+8(r1) + lfd fp1,FPR_RTN+0(r1) + lfd fp2,FPR_RTN+8(r1) cmpdi cr0,r12,0 - la r10,VR_RTN(r1) - lfd fp3,FPR_PARMS+16(r1) - lfd fp4,FPR_PARMS+24(r1) + la r11,VR_RTN(r1) + lfd fp3,FPR_RTN+16(r1) + lfd fp4,FPR_RTN+24(r1) +#if _CALL_ELF == 2 + la r12,VR_RTN+16(r1) + lfd fp5,FPR_RTN+32(r1) + lfd fp6,FPR_RTN+40(r1) + li r30,32 + li r31,64 + lfd fp7,FPR_RTN+48(r1) + lfd fp8,FPR_RTN+56(r1) + lfd fp9,FPR_RTN+64(r1) + lfd fp10,FPR_RTN+72(r1) +#endif beq L(pltexitreturn) - lvx v2,0,r10 + lvx v2,0,r11 +#if _CALL_ELF == 2 + lvx v3,0,r12 + lvx v4,r30,r11 + lvx v5,r30,r12 + addi r30,r30,64 + lvx v6,r31,r11 + lvx v7,r31,r12 + lvx v8,r30,r11 + lvx v9,r30,r12 +#endif L(pltexitreturn): ld r0,FRAME_SIZE+FRAME_LR_SAVE(r1) - ld r31,584(r1) - ld r30,576(r1) + ld r31,FRAME_SIZE-8(r1) + ld r30,FRAME_SIZE-16(r1) mtlr r0 ld r1,0(r1) blr |