diff options
| author | Erik de Castro Lopo <erikd@mega-nerd.com> | 2014-01-07 21:25:03 +1100 |
|---|---|---|
| committer | Erik de Castro Lopo <erikd@mega-nerd.com> | 2014-01-07 21:27:09 +1100 |
| commit | 8e4a45ac863a63e4f6256be487e71e852f22fbba (patch) | |
| tree | 97311002e67fd3b82f25679cb94dee66d2332c9b | |
| parent | 6cd8b42438e6959ae60ae3bea84b27d2553f50c9 (diff) | |
| download | flac-8e4a45ac863a63e4f6256be487e71e852f22fbba.tar.gz | |
libFLAC/ia32/lpc_asm.nasm : Match calls and returns.
According to Agner Fog, "...you must make sure that all calls
are matched with returns. Never jump out of a subroutine without
a return and never use a return as an indirect jump."
(see paragraph 3.15 in microarchitecture.pdf and
examples 3.5a and 3.5b in optimizing_assembly.pdf)
Patch-from: lvqcl <lvqcl.mail@gmail.com>
| -rw-r--r-- | src/libFLAC/ia32/lpc_asm.nasm | 24 |
1 files changed, 16 insertions, 8 deletions
diff --git a/src/libFLAC/ia32/lpc_asm.nasm b/src/libFLAC/ia32/lpc_asm.nasm index 66d07e2d..2b7a8019 100644 --- a/src/libFLAC/ia32/lpc_asm.nasm +++ b/src/libFLAC/ia32/lpc_asm.nasm @@ -115,9 +115,8 @@ cident FLAC__lpc_compute_autocorrelation_asm_ia32 lea edx, [eax + eax*2] neg edx lea edx, [eax + edx*4 + .jumper1_0 - .get_eip1] - call .get_eip1 + call .mov_eip_to_ebx .get_eip1: - pop ebx add edx, ebx inc edx ; compensate for the shorter opcode on the last iteration inc edx ; compensate for the shorter opcode on the last iteration @@ -128,6 +127,10 @@ cident FLAC__lpc_compute_autocorrelation_asm_ia32 .loop1_start: jmp edx +.mov_eip_to_ebx: + mov ebx, [esp] + ret + fld st0 ; ST = d d fmul dword [esi + (32*4)] ; ST = d*data[sample+32] d WATCHOUT: not a byte displacement here! fadd dword [edi + (32*4)] ; ST = autoc[32]+d*data[sample+32] d WATCHOUT: not a byte displacement here! @@ -285,9 +288,8 @@ cident FLAC__lpc_compute_autocorrelation_asm_ia32 lea edx, [eax + eax*2] neg edx lea edx, [eax + edx*4 + .jumper2_0 - .get_eip2] - call .get_eip2 + call .mov_eip_to_ebx .get_eip2: - pop ebx add edx, ebx inc edx ; compensate for the shorter opcode on the last iteration inc edx ; compensate for the shorter opcode on the last iteration @@ -919,13 +921,16 @@ cident FLAC__lpc_compute_residual_from_qlp_coefficients_asm_ia32 jmp .end +.mov_eip_to_eax: + mov eax, [esp] + ret + .i_32: sub edi, esi neg eax lea edx, [eax + eax * 8 + .jumper_0 - .get_eip0] - call .get_eip0 + call .mov_eip_to_eax .get_eip0: - pop eax add edx, eax inc edx mov eax, [esp + 28] ; eax = qlp_coeff[] @@ -1323,13 +1328,16 @@ cident FLAC__lpc_restore_signal_asm_ia32 jmp .end +.mov_eip_to_eax: + mov eax, [esp] + ret + .x87_32: sub esi, edi neg eax lea edx, [eax + eax * 8 + .jumper_0 - .get_eip0] - call .get_eip0 + call .mov_eip_to_eax .get_eip0: - pop eax add edx, eax inc edx ; compensate for the shorter opcode on the last iteration mov eax, [esp + 28] ; eax = qlp_coeff[] |