diff options
author | Rich Felker <dalias@aerifal.cx> | 2016-01-21 19:28:15 +0000 |
---|---|---|
committer | Rich Felker <dalias@aerifal.cx> | 2016-01-21 19:43:04 +0000 |
commit | 61b1e75f7d8004461f2e18f171c26c2f545eed32 (patch) | |
tree | 7a88f8d51ee3f049745b147b7f36fc8ee22e0e62 /src/thread/sh | |
parent | 1315596b510189b5159e742110b504177bdd4932 (diff) | |
download | musl-61b1e75f7d8004461f2e18f171c26c2f545eed32.tar.gz musl-61b1e75f7d8004461f2e18f171c26c2f545eed32.tar.bz2 musl-61b1e75f7d8004461f2e18f171c26c2f545eed32.tar.xz musl-61b1e75f7d8004461f2e18f171c26c2f545eed32.zip |
overhaul sh atomics for new atomics framework, add j-core cas.l backend
sh needs runtime-selected atomic backends since there are a number of
supported models that use non-forwards-compatible (non-smp-compatible)
atomic mechanisms. previously, the code paths for this were highly
inefficient since they involved C function calls with multiple
branches in the callee and heavy spills in the caller. the new code
performs calls the runtime-selected asm fragment from inline asm with
extremely minimal clobbers, rather than using a function call.
for the sh4a case where the atomic mechanism is known and there is no
forward-compatibility issue, the movli.l and movco.l instructions are
provided as a_ll and a_sc, allowing the new shared atomic.h to
generate efficient inline versions of all the basic atomic operations
without needing a cas loop.
Diffstat (limited to 'src/thread/sh')
-rw-r--r-- | src/thread/sh/__set_thread_area.c | 40 | ||||
-rw-r--r-- | src/thread/sh/__set_thread_area.s | 0 | ||||
-rw-r--r-- | src/thread/sh/atomics.s | 65 |
3 files changed, 105 insertions, 0 deletions
diff --git a/src/thread/sh/__set_thread_area.c b/src/thread/sh/__set_thread_area.c new file mode 100644 index 00000000..9c47f78d --- /dev/null +++ b/src/thread/sh/__set_thread_area.c @@ -0,0 +1,40 @@ +#include "pthread_impl.h" +#include "libc.h" +#include <elf.h> + +/* Also perform sh-specific init */ + +#define CPU_HAS_LLSC 0x0040 +#define CPU_HAS_CAS_L 0x0400 + +__attribute__((__visibility__("hidden"))) +extern const char __sh_cas_gusa[], __sh_cas_llsc[], __sh_cas_imask[], __sh_cas_cas_l[]; + +__attribute__((__visibility__("hidden"))) +const void *__sh_cas_ptr; + +__attribute__((__visibility__("hidden"))) +unsigned __sh_nommu; + +int __set_thread_area(void *p) +{ + size_t *aux; + __asm__ __volatile__ ( "ldc %0, gbr" : : "r"(p) : "memory" ); +#ifndef __SH4A__ + __sh_cas_ptr = __sh_cas_gusa; +#if !defined(__SH3__) && !defined(__SH4__) + for (aux=libc.auxv; *aux; aux+=2) { + if (*aux != AT_PLATFORM) continue; + const char *s = (void *)aux[1]; + if (s[0]!='s' || s[1]!='h' || s[2]!='2' || s[3]-'0'<10u) break; + __sh_cas_ptr = __sh_cas_imask; + __sh_nommu = 1; + } +#endif + if (__hwcap & CPU_HAS_CAS_L) + __sh_cas_ptr = __sh_cas_cas_l; + else if (__hwcap & CPU_HAS_LLSC) + __sh_cas_ptr = __sh_cas_llsc; +#endif + return 0; +} diff --git a/src/thread/sh/__set_thread_area.s b/src/thread/sh/__set_thread_area.s deleted file mode 100644 index e69de29b..00000000 --- a/src/thread/sh/__set_thread_area.s +++ /dev/null diff --git a/src/thread/sh/atomics.s b/src/thread/sh/atomics.s new file mode 100644 index 00000000..3b58cccc --- /dev/null +++ b/src/thread/sh/atomics.s @@ -0,0 +1,65 @@ +/* Contract for all versions is same as cas.l r2,r3,@r0 + * pr and r1 are also clobbered (by jsr & r1 as temp). + * r0,r2,r4-r15 must be preserved. + * r3 contains result (==r2 iff cas succeeded). */ + + .align 2 +.global __sh_cas_gusa +.hidden __sh_cas_gusa +__sh_cas_gusa: + mov.l r5,@-r15 + mov.l r4,@-r15 + mov r0,r4 + mova 1f,r0 + mov r15,r1 + mov #(0f-1f),r15 +0: mov.l @r4,r5 + cmp/eq r5,r2 + bf 1f + mov.l r3,@r4 +1: mov r1,r15 + mov r5,r3 + mov r4,r0 + mov.l @r15+,r4 + rts + mov.l @r15+,r5 + +.global __sh_cas_llsc +.hidden __sh_cas_llsc +__sh_cas_llsc: + mov r0,r1 + synco +0: movli.l @r1,r0 + cmp/eq r0,r2 + bf 1f + mov r3,r0 + movco.l r0,@r1 + bf 0b + mov r2,r0 +1: synco + mov r0,r3 + rts + mov r1,r0 + +.global __sh_cas_imask +.hidden __sh_cas_imask +__sh_cas_imask: + mov r0,r1 + stc sr,r0 + mov.l r0,@-r15 + or #0xf0,r0 + ldc r0,sr + mov.l @r1,r0 + cmp/eq r0,r2 + bf 1f + mov.l r3,@r1 +1: ldc.l @r15+,sr + mov r0,r3 + rts + mov r1,r0 + +.global __sh_cas_cas_l +.hidden __sh_cas_cas_l +__sh_cas_cas_l: + rts + .word 0x2323 /* cas.l r2,r3,@r0 */ |