Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | arm/neon: timestamp counter (not for master)neon | David Schleef | 2010-07-09 | 1 | -0/+18 |
| | | | | | Add timestamp counter for Cortex-A8. Requires a kernel patch to use. | ||||
* | memcpy_speed: preload distance check (not for master) | David Schleef | 2010-07-09 | 1 | -1/+161 |
| | |||||
* | neon: preload hinting, not for master | David Schleef | 2010-07-09 | 1 | -0/+15 |
| | |||||
* | neon: Add load_constant (not for master) | David Schleef | 2010-07-09 | 1 | -1/+20 |
| | |||||
* | add performance numbers. not for master | David Schleef | 2010-07-09 | 1 | -0/+40 |
| | |||||
* | neon: Fix aligned load | David Schleef | 2010-07-09 | 1 | -1/+1 |
| | |||||
* | memcpy_speed: inline orc_memcpy | David Schleef | 2010-07-09 | 1 | -1/+28 |
| | |||||
* | memcpy_speed: flush/populate cache for hot/cold testing | David Schleef | 2010-07-09 | 1 | -0/+42 |
| | |||||
* | neon: Add short-array unaligned speedup | David Schleef | 2010-07-09 | 2 | -2/+66 |
| | |||||
* | neon: Add load/store of 256-bit vectors | David Schleef | 2010-07-09 | 1 | -2/+61 |
| | |||||
* | neon: Only preload for each cache line | David Schleef | 2010-07-09 | 1 | -9/+11 |
| | |||||
* | neon: Add handling of different size regions | David Schleef | 2010-07-09 | 2 | -21/+99 |
| | |||||
* | arm/neon: use subs for loop counter | David Schleef | 2010-07-09 | 4 | -14/+10 |
| | |||||
* | neon: add preload | David Schleef | 2010-07-09 | 3 | -1/+30 |
| | |||||
* | neon: add loop unrolling | David Schleef | 2010-07-09 | 1 | -1/+6 |
| | |||||
* | neon: Fix rules for quad usage | David Schleef | 2010-07-09 | 2 | -64/+142 |
| | |||||
* | memcpy_speed: take unalignment parameter on cmd line | David Schleef | 2010-07-09 | 1 | -6/+14 |
| | |||||
* | neon: Fix store of l vectors | David Schleef | 2010-07-09 | 1 | -2/+2 |
| | |||||
* | neon: Add quad load/store | David Schleef | 2010-07-09 | 2 | -14/+103 |
| | |||||
* | neon: Use faster opcodes for aligned store | David Schleef | 2010-07-09 | 1 | -6/+6 |
| | |||||
* | Add simple constant folding | David Schleef | 2010-07-09 | 1 | -1/+10 |
| | |||||
* | neon: Use unaligned ops instead of mask register | David Schleef | 2010-07-09 | 2 | -3/+42 |
| | |||||
* | Improve compile debugging | David Schleef | 2010-07-09 | 1 | -1/+4 |
| | |||||
* | test: remove adjustment of performance number | David Schleef | 2010-07-09 | 1 | -1/+1 |
| | |||||
* | test: switch neon test back to neon | David Schleef | 2010-07-09 | 1 | -1/+1 |
| | |||||
* | win32: Fix initialization of once mutex | David Schleef | 2010-07-09 | 1 | -1/+9 |
| | |||||
* | arm: Fix register moving | David Schleef | 2010-07-05 | 2 | -1/+2 |
| | |||||
* | arm: fix some compile mismatches | David Schleef | 2010-07-05 | 2 | -12/+20 |
| | |||||
* | arm: Fix a bunch of rules | David Schleef | 2010-07-01 | 1 | -25/+32 |
| | |||||
* | arm: fix compile error | David Schleef | 2010-07-01 | 1 | -1/+1 |
| | |||||
* | arm: mark rules known to be buggy | David Schleef | 2010-06-30 | 1 | -44/+49 |
| | |||||
* | arm: Implement 2d support | David Schleef | 2010-06-30 | 1 | -5/+92 |
| | |||||
* | arm: implement load/store of b, l | David Schleef | 2010-06-30 | 3 | -6/+83 |
| | |||||
* | arm: arm backend code gen fixes | David Schleef | 2010-06-30 | 4 | -24/+63 |
| | |||||
* | neon: check for NEON support on CPU | David Schleef | 2010-06-30 | 4 | -4/+65 |
| | |||||
* | sse: handle NANs and denormals | David Schleef | 2010-06-29 | 7 | -3/+114 |
| | | | | | | Set the DAZ and FTZ flags in the MXCSR, to set proper denormal behavior. Implement NAN handling for maxf and minf. | ||||
* | opcodes: Make flushing denormals the default behavior | David Schleef | 2010-06-29 | 1 | -4/+12 |
| | | | | | | | | | | Flushing denormals is a better default behavior for SIMD engines, since they tend handle denormals slowly (SSE, Altivec) or not at all (NEON). Unfortunately, that behavior is expensive to emulate on IEEE-compliant FPUs, since you need to check for denormal before and after every operation. At this point, the emulator flushes denormals and C code does not. | ||||
* | x86: Fix register popping for x86_64 | David Schleef | 2010-06-28 | 1 | -1/+1 |
| | |||||
* | opcodes: Add code for handling denormals | David Schleef | 2010-06-28 | 1 | -7/+15 |
| | |||||
* | opcodes: check for NaN in maxf/minf | David Schleef | 2010-06-28 | 1 | -2/+2 |
| | |||||
* | tests: better float comparison | David Schleef | 2010-06-28 | 2 | -7/+35 |
| | | | | Allow denormals to be flushed to 0. Check for NaNs. | ||||
* | neon: Fix accumulator opcodes | David Schleef | 2010-06-28 | 2 | -24/+97 |
| | |||||
* | neon: implement 2d arrays | David Schleef | 2010-06-28 | 1 | -2/+63 |
| | |||||
* | Fix segfaults if no targets are available | David Schleef | 2010-06-28 | 1 | -1/+12 |
| | |||||
* | Add support for OpenBSD in CPU detection | David Schleef | 2010-06-28 | 2 | -6/+32 |
| | | | | From Brad <brad@comstyle.com>. | ||||
* | tools: add orc-bugreport | David Schleef | 2010-06-26 | 2 | -1/+516 |
| | |||||
* | orcc: Add --quiet and --benchmark parameters to generated test programs | Sebastian Dröge | 2010-06-26 | 1 | -8/+51 |
| | |||||
* | orcc: Improve test program generation | Sebastian Dröge | 2010-06-26 | 1 | -9/+8 |
| | | | | Output something if a check has passed instead of being completely silent | ||||
* | target: add function to return flag names | David Schleef | 2010-06-26 | 4 | -3/+51 |
| | |||||
* | arm: disable unused CPU detection | David Schleef | 2010-06-24 | 1 | -0/+4 |
| |