Add support for ARM NEON targets.

Initial support for ARM NEON on Cortex-A9 and A15 CPUs.  All but ~10 tests
pass, and all examples compile and run correctly.  Most of the examples
show a ~2x speedup on a single A15 core versus scalar code.

Current open issues/TODOs
- Code quality looks decent, but hasn't been carefully examined.  Known
  issues/opportunities for improvement include:
  - fp32 vector divide is done as a series of scalar divides rather than
    a vector divide (which I believe exists, but I may be mistaken.)
    This is particularly harmful to examples/rt, which only runs ~1.5x
    faster with ispc, likely due to long chains of scalar divides.
  - The compiler isn't generating a vmin.f32 for e.g. the final scalar
    min in reduce_min(); instead it's generating a compare and then a
    select instruction (and similarly elsewhere).
  - There are some additional FIXMEs in builtins/target-neon.ll that
    include both a few pieces of missing functionality (e.g. rounding
    doubles) as well as places that deserve attention for possible
    code quality improvements.

- Currently only the "cortex-a9" and "cortex-15" CPU targets are
  supported; LLVM supports many other ARM CPUs and ispc should provide
  access to all of the ones that have NEON support (and aren't too
  obscure.)

- ~5 of the reduce-* tests hit an assertion inside LLVM (unfortunately
   only when the compiler runs on an ARM host, though).

- The Windows build hasn't been tested (though I've tried to update
  ispc.vcxproj appropriately).  It may just work, but will more likely
  have various small issues.)

- Anything related to 64-bit ARM has seen no attention.
This commit is contained in:
Matt Pharr
2013-07-19 11:06:11 -07:00
parent b007bba59f
commit d7b0c5794e
22 changed files with 914 additions and 67 deletions

View File

@@ -33,33 +33,47 @@
#include <stdint.h>
#ifdef __arm__
#include <sys/time.h>
// There's no easy way to get a hardware clock counter on ARM, so instead
// we'll pretend it's a 1GHz processor and then compute pretend cycles
// based on elapsed time from gettimeofday().
__inline__ uint64_t rdtsc() {
static bool first = true;
static struct timeval tv_start;
if (first) {
gettimeofday(&tv_start, NULL);
first = false;
return 0;
}
struct timeval tv;
gettimeofday(&tv, NULL);
tv.tv_sec -= tv_start.tv_sec;
tv.tv_usec -= tv_start.tv_usec;
return (1000000ull * tv.tv_sec + tv.tv_usec) * 1000ull;
}
#else // __arm__
#ifdef WIN32
#include <windows.h>
#define rdtsc __rdtsc
#else
#ifdef __cplusplus
extern "C" {
#endif /* __cplusplus */
__inline__ uint64_t rdtsc() {
uint32_t low, high;
#else // WIN32
__inline__ uint64_t rdtsc() {
uint32_t low, high;
#ifdef __x86_64
__asm__ __volatile__ (
"xorl %%eax,%%eax \n cpuid"
::: "%rax", "%rbx", "%rcx", "%rdx" );
__asm__ __volatile__ ("xorl %%eax,%%eax \n cpuid"
::: "%rax", "%rbx", "%rcx", "%rdx" );
#else
__asm__ __volatile__ (
"xorl %%eax,%%eax \n cpuid"
::: "%eax", "%ebx", "%ecx", "%edx" );
__asm__ __volatile__ ("xorl %%eax,%%eax \n cpuid"
::: "%eax", "%ebx", "%ecx", "%edx" );
#endif
__asm__ __volatile__ (
"rdtsc" : "=a" (low), "=d" (high));
return (uint64_t)high << 32 | low;
}
#ifdef __cplusplus
__asm__ __volatile__ ("rdtsc" : "=a" (low), "=d" (high));
return (uint64_t)high << 32 | low;
}
#endif /* __cplusplus */
#endif
#endif // !WIN32
#endif // !__arm__
static uint64_t start, end;