Remove redundant space in effective address

To have effective address of vmovapd instruction recognized and
translated to Intel format by the x86_64 assembler translator.

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/27632)

diff --git a/crypto/bn/asm/rsaz-2k-avxifma.pl b/crypto/bn/asm/rsaz-2k-avxifma.pl
index 65710eac3ec4c58402c6bf457ffc0bb55c579d5f..52af336f0ecdbcee535a12e58b5a5df69bf7af9f 100644 (file)
--- a/crypto/bn/asm/rsaz-2k-avxifma.pl
+++ b/crypto/bn/asm/rsaz-2k-avxifma.pl
@@ -296,34 +296,34 @@ $code.=<<___;
     and \$0xf, %r14
     vpsubq  .Lmask52x4(%rip), $_R0,  $T0
     shl \$5, %r14
-    vmovapd (%rdx, %r14), $T1
+    vmovapd (%rdx,%r14), $T1
     vblendvpd $T1, $T0, $_R0, $_R0
 
     shr \$4, %r13b
     and \$0xf, %r13
     vpsubq  .Lmask52x4(%rip), $_R0h,  $T0
     shl \$5, %r13
-    vmovapd (%rdx, %r13), $T1
+    vmovapd (%rdx,%r13), $T1
     vblendvpd $T1, $T0, $_R0h, $_R0h
 
     mov %r12b, %r11b
     and \$0xf, %r12
     vpsubq  .Lmask52x4(%rip), $_R1,  $T0
     shl \$5, %r12
-    vmovapd (%rdx, %r12), $T1
+    vmovapd (%rdx,%r12), $T1
     vblendvpd $T1, $T0, $_R1, $_R1
 
     shr \$4, %r11b
     and \$0xf, %r11
     vpsubq  .Lmask52x4(%rip), $_R1h,  $T0
     shl \$5, %r11
-    vmovapd (%rdx, %r11), $T1
+    vmovapd (%rdx,%r11), $T1
     vblendvpd $T1, $T0, $_R1h, $_R1h
 
     and \$0xf, %r10
     vpsubq  .Lmask52x4(%rip), $_R2,  $T0
     shl \$5, %r10
-    vmovapd (%rdx, %r10), $T1
+    vmovapd (%rdx,%r10), $T1
     vblendvpd $T1, $T0, $_R2, $_R2
 
     # Add carries according to the obtained mask

diff --git a/crypto/bn/asm/rsaz-3k-avxifma.pl b/crypto/bn/asm/rsaz-3k-avxifma.pl
index 64c00c10e325c69d4735044be43ef7fff237d0db..36c3c28813006094a07b3a9c9a66c94fe613ec71 100644 (file)
--- a/crypto/bn/asm/rsaz-3k-avxifma.pl
+++ b/crypto/bn/asm/rsaz-3k-avxifma.pl
@@ -355,56 +355,56 @@ $code.=<<___;
     and \$0xf, %r14
     vpsubq  .Lmask52x4(%rip), $_R0,  $T0
     shl \$5, %r14
-    vmovapd (%rdx, %r14), $T1
+    vmovapd (%rdx,%r14), $T1
     vblendvpd $T1, $T0, $_R0, $_R0
 
     shr \$4, %r10b
     and \$0xf, %r10
     vpsubq  .Lmask52x4(%rip), $_R0h,  $T0
     shl \$5, %r10
-    vmovapd (%rdx, %r10), $T1
+    vmovapd (%rdx,%r10), $T1
     vblendvpd $T1, $T0, $_R0h, $_R0h
 
     mov %r13b, %r10b
     and \$0xf, %r13
     vpsubq  .Lmask52x4(%rip), $_R1,  $T0
     shl \$5, %r13
-    vmovapd (%rdx, %r13), $T1
+    vmovapd (%rdx,%r13), $T1
     vblendvpd $T1, $T0, $_R1, $_R1
 
     shr \$4, %r10b
     and \$0xf, %r10
     vpsubq  .Lmask52x4(%rip), $_R1h,  $T0
     shl \$5, %r10
-    vmovapd (%rdx, %r10), $T1
+    vmovapd (%rdx,%r10), $T1
     vblendvpd $T1, $T0, $_R1h, $_R1h
 
     mov %r12b, %r10b
     and \$0xf, %r12
     vpsubq  .Lmask52x4(%rip), $_R2,  $T0
     shl \$5, %r12
-    vmovapd (%rdx, %r12), $T1
+    vmovapd (%rdx,%r12), $T1
     vblendvpd $T1, $T0, $_R2, $_R2
 
     shr \$4, %r10b
     and \$0xf, %r10
     vpsubq  .Lmask52x4(%rip), $_R2h,  $T0
     shl \$5, %r10
-    vmovapd (%rdx, %r10), $T1
+    vmovapd (%rdx,%r10), $T1
     vblendvpd $T1, $T0, $_R2h, $_R2h
 
     mov %r11b, %r10b
     and \$0xf, %r11
     vpsubq  .Lmask52x4(%rip), $_R3,  $T0
     shl \$5, %r11
-    vmovapd (%rdx, %r11), $T1
+    vmovapd (%rdx,%r11), $T1
     vblendvpd $T1, $T0, $_R3, $_R3
 
     shr \$4, %r10b
     and \$0xf, %r10
     vpsubq  .Lmask52x4(%rip), $_R3h,  $T0
     shl \$5, %r10
-    vmovapd (%rdx, %r10), $T1
+    vmovapd (%rdx,%r10), $T1
     vblendvpd $T1, $T0, $_R3h, $_R3h
 
     vpand  .Lmask52x4(%rip), $_R0,  $_R0

diff --git a/crypto/bn/asm/rsaz-4k-avxifma.pl b/crypto/bn/asm/rsaz-4k-avxifma.pl
index 5de6a4c4c2b2fb6c8aaf0aa04089a0f3b933173e..f472ec2c18acb6b0fed596e75c311fbba6633792 100644 (file)
--- a/crypto/bn/asm/rsaz-4k-avxifma.pl
+++ b/crypto/bn/asm/rsaz-4k-avxifma.pl
@@ -424,70 +424,70 @@ $code.=<<___;
     and \$0xf, %r14
     vpsubq  .Lmask52x4(%rip), $_R0,  $tmp
     shl \$5, %r14
-    vmovapd (%r8, %r14), $tmp2
+    vmovapd (%r8,%r14), $tmp2
     vblendvpd $tmp2, $tmp, $_R0, $_R0
 
     shr \$4, %r9b
     and \$0xf, %r9
     vpsubq  .Lmask52x4(%rip), $_R0h,  $tmp
     shl \$5, %r9
-    vmovapd (%r8, %r9), $tmp2
+    vmovapd (%r8,%r9), $tmp2
     vblendvpd $tmp2, $tmp, $_R0h, $_R0h
 
     movb    %r13b,%r9b
     and \$0xf, %r13
     vpsubq  .Lmask52x4(%rip), $_R1,  $tmp
     shl \$5, %r13
-    vmovapd (%r8, %r13), $tmp2
+    vmovapd (%r8,%r13), $tmp2
     vblendvpd $tmp2, $tmp, $_R1, $_R1
 
     shr \$4, %r9b
     and \$0xf, %r9
     vpsubq  .Lmask52x4(%rip), $_R1h,  $tmp
     shl \$5, %r9
-    vmovapd (%r8, %r9), $tmp2
+    vmovapd (%r8,%r9), $tmp2
     vblendvpd $tmp2, $tmp, $_R1h, $_R1h
 
     movb    %r12b,%r9b
     and \$0xf, %r12
     vpsubq  .Lmask52x4(%rip), $_R2,  $tmp
     shl \$5, %r12
-    vmovapd (%r8, %r12), $tmp2
+    vmovapd (%r8,%r12), $tmp2
     vblendvpd $tmp2, $tmp, $_R2, $_R2
 
     shr \$4, %r9b
     and \$0xf, %r9
     vpsubq  .Lmask52x4(%rip), $_R2h,  $tmp
     shl \$5, %r9
-    vmovapd (%r8, %r9), $tmp2
+    vmovapd (%r8,%r9), $tmp2
     vblendvpd $tmp2, $tmp, $_R2h, $_R2h
 
     movb    %r11b,%r9b
     and \$0xf, %r11
     vpsubq  .Lmask52x4(%rip), $_R3,  $tmp
     shl \$5, %r11
-    vmovapd (%r8, %r11), $tmp2
+    vmovapd (%r8,%r11), $tmp2
     vblendvpd $tmp2, $tmp, $_R3, $_R3
 
     shr \$4, %r9b
     and \$0xf, %r9
     vpsubq  .Lmask52x4(%rip), $_R3h,  $tmp
     shl \$5, %r9
-    vmovapd (%r8, %r9), $tmp2
+    vmovapd (%r8,%r9), $tmp2
     vblendvpd $tmp2, $tmp, $_R3h, $_R3h
 
     movb    %r10b,%r9b
     and \$0xf, %r10
     vpsubq  .Lmask52x4(%rip), $_R4,  $tmp
     shl \$5, %r10
-    vmovapd (%r8, %r10), $tmp2
+    vmovapd (%r8,%r10), $tmp2
     vblendvpd $tmp2, $tmp, $_R4, $_R4
 
     shr \$4, %r9b
     and \$0xf, %r9
     vpsubq  .Lmask52x4(%rip), $_R4h,  $tmp
     shl \$5, %r9
-    vmovapd (%r8, %r9), $tmp2
+    vmovapd (%r8,%r9), $tmp2
     vblendvpd $tmp2, $tmp, $_R4h, $_R4h
 
     pop %r8