So far, AV_READ_TIME would return the cycle counter. This posed two
problems:
1) On recent systems, it would just raise an illegal instruction
exception. Indeed RDCYCLE is blocked in user space to ward off some
side channel attacks. In particular, this would cause the random
number generator to crash.
2) It does not match the x86 behaviour and the apparent original intent
of AV_READ_TIME in the functional code base (outside test cases).
So this replaces the cycle counter with the time counter. The unit is
a platform-dependent constant fraction of time, and the value should be
stable across harts (RISC-V lingo for physical CPU thread).
---
libavutil/riscv/timer.h | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/libavutil/riscv/timer.h b/libavutil/riscv/timer.h
index a34157a566..c2465a4524 100644
--- a/libavutil/riscv/timer.h
+++ b/libavutil/riscv/timer.h
@@ -24,21 +24,21 @@
#if HAVE_INLINE_ASM
#include <stdint.h>
-static inline uint64_t rdcycle64(void)
+static inline uint64_t ff_read_time(void)
{
#if (__riscv_xlen >= 64)
uintptr_t cycles;
- __asm__ volatile ("rdcycle %0" : "=r"(cycles));
+ __asm__ volatile ("rdtime %0" : "=r" (cycles));
#else
uint64_t cycles;
uint32_t hi, lo, check;
__asm__ volatile (
- "1: rdcycleh %0\n"
- " rdcycle %1\n"
- " rdcycleh %2\n"
+ "1: rdtimeh %0\n"
+ " rdtime %1\n"
+ " rdtimeh %2\n"
" bne %0, %2, 1b\n" : "=r" (hi), "=r" (lo), "=r" (check));
cycles = (((uint64_t)hi) << 32) | lo;
@@ -47,7 +47,7 @@ static inline uint64_t rdcycle64(void)
return cycles;
}
-#define AV_READ_TIME rdcycle64
+#define AV_READ_TIME ff_read_time
#endif
#endif /* AVUTIL_RISCV_TIMER_H */
--
2.40.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
