summaryrefslogtreecommitdiffstats
path: root/include/asm-mips/bitops.h
diff options
context:
space:
mode:
authorRalf Baechle <ralf@linux-mips.org>2006-09-28 01:45:21 +0100
committerRalf Baechle <ralf@linux-mips.org>2007-02-13 22:40:50 +0000
commitf65e4fa8e0c6022ad58dc88d1b11b12589ed7f9f (patch)
tree2405e012e079693e0fcfde9ff981c549d6c68a21 /include/asm-mips/bitops.h
parent509cb37e173d4e39cec47238397e91b718730794 (diff)
downloadtalos-obmc-linux-f65e4fa8e0c6022ad58dc88d1b11b12589ed7f9f.tar.gz
talos-obmc-linux-f65e4fa8e0c6022ad58dc88d1b11b12589ed7f9f.zip
[MIPS] Improve branch prediction in ll/sc atomic operations.
Now that finally all supported versions of binutils have functioning support for .subsection use .subsection to tweak the branch prediction I did not modify the R10000 errata variants because it seems unclear if this will invalidate the workaround which actually relies on the cheesy prediction of branch likely to cause a misspredict if the sc was successful. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Diffstat (limited to 'include/asm-mips/bitops.h')
-rw-r--r--include/asm-mips/bitops.h33
1 files changed, 27 insertions, 6 deletions
diff --git a/include/asm-mips/bitops.h b/include/asm-mips/bitops.h
index 06445de1324b..06c08228a525 100644
--- a/include/asm-mips/bitops.h
+++ b/include/asm-mips/bitops.h
@@ -68,7 +68,10 @@ static inline void set_bit(unsigned long nr, volatile unsigned long *addr)
"1: " __LL "%0, %1 # set_bit \n"
" or %0, %2 \n"
" " __SC "%0, %1 \n"
- " beqz %0, 1b \n"
+ " beqz %0, 2f \n"
+ " .subsection 2 \n"
+ "2: b 1b \n"
+ " .previous \n"
" .set mips0 \n"
: "=&r" (temp), "=m" (*m)
: "ir" (1UL << (nr & SZLONG_MASK)), "m" (*m));
@@ -116,7 +119,10 @@ static inline void clear_bit(unsigned long nr, volatile unsigned long *addr)
"1: " __LL "%0, %1 # clear_bit \n"
" and %0, %2 \n"
" " __SC "%0, %1 \n"
- " beqz %0, 1b \n"
+ " beqz %0, 2f \n"
+ " .subsection 2 \n"
+ "2: b 1b \n"
+ " .previous \n"
" .set mips0 \n"
: "=&r" (temp), "=m" (*m)
: "ir" (~(1UL << (nr & SZLONG_MASK))), "m" (*m));
@@ -166,7 +172,10 @@ static inline void change_bit(unsigned long nr, volatile unsigned long *addr)
"1: " __LL "%0, %1 # change_bit \n"
" xor %0, %2 \n"
" " __SC "%0, %1 \n"
- " beqz %0, 1b \n"
+ " beqz %0, 2f \n"
+ " .subsection 2 \n"
+ "2: b 1b \n"
+ " .previous \n"
" .set mips0 \n"
: "=&r" (temp), "=m" (*m)
: "ir" (1UL << (nr & SZLONG_MASK)), "m" (*m));
@@ -222,8 +231,12 @@ static inline int test_and_set_bit(unsigned long nr,
"1: " __LL "%0, %1 # test_and_set_bit \n"
" or %2, %0, %3 \n"
" " __SC "%2, %1 \n"
- " beqz %2, 1b \n"
+ " beqz %2, 2f \n"
" and %2, %0, %3 \n"
+ " .subsection 2 \n"
+ "2: b 1b \n"
+ " nop \n"
+ " .previous \n"
" .set pop \n"
: "=&r" (temp), "=m" (*m), "=&r" (res)
: "r" (1UL << (nr & SZLONG_MASK)), "m" (*m)
@@ -290,8 +303,12 @@ static inline int test_and_clear_bit(unsigned long nr,
" or %2, %0, %3 \n"
" xor %2, %3 \n"
" " __SC "%2, %1 \n"
- " beqz %2, 1b \n"
+ " beqz %2, 2f \n"
" and %2, %0, %3 \n"
+ " .subsection 2 \n"
+ "2: b 1b \n"
+ " nop \n"
+ " .previous \n"
" .set pop \n"
: "=&r" (temp), "=m" (*m), "=&r" (res)
: "r" (1UL << (nr & SZLONG_MASK)), "m" (*m)
@@ -356,8 +373,12 @@ static inline int test_and_change_bit(unsigned long nr,
"1: " __LL "%0, %1 # test_and_change_bit \n"
" xor %2, %0, %3 \n"
" " __SC "\t%2, %1 \n"
- " beqz %2, 1b \n"
+ " beqz %2, 2f \n"
" and %2, %0, %3 \n"
+ " .subsection 2 \n"
+ "2: b 1b \n"
+ " nop \n"
+ " .previous \n"
" .set pop \n"
: "=&r" (temp), "=m" (*m), "=&r" (res)
: "r" (1UL << (nr & SZLONG_MASK)), "m" (*m)
OpenPOWER on IntegriCloud