diff options
| author | Eric Anholt <[email protected]> | 2015-03-29 21:21:10 -0700 |
|---|---|---|
| committer | Eric Anholt <[email protected]> | 2015-03-30 09:23:39 -0700 |
| commit | 5df8bf86fe40ae95ad3888cb167ce80c710af227 (patch) | |
| tree | 8164de0719c15a4d60b6a10d92b1544ee98b85b5 | |
| parent | 18004c338f6be8af2e36d2f54972c60136229aeb (diff) | |
vc4: Make integer multiply use 24 bits for the low parts.
The hardware uses the low 24 bits in integer multiplies, so we can have
fewer high bits (and so probably drop them more frequently).
| -rw-r--r-- | src/gallium/drivers/vc4/vc4_program.c | 10 |
1 files changed, 5 insertions, 5 deletions
diff --git a/src/gallium/drivers/vc4/vc4_program.c b/src/gallium/drivers/vc4/vc4_program.c index 56a3a96c1b5..49b94666b88 100644 --- a/src/gallium/drivers/vc4/vc4_program.c +++ b/src/gallium/drivers/vc4/vc4_program.c @@ -367,13 +367,13 @@ tgsi_to_qir_umul(struct vc4_compile *c, enum qop op, struct qreg *src, int i) { struct qreg src0_hi = qir_SHR(c, src[0 * 4 + i], - qir_uniform_ui(c, 16)); + qir_uniform_ui(c, 24)); struct qreg src0_lo = qir_AND(c, src[0 * 4 + i], - qir_uniform_ui(c, 0xffff)); + qir_uniform_ui(c, 0xffffff)); struct qreg src1_hi = qir_SHR(c, src[1 * 4 + i], - qir_uniform_ui(c, 16)); + qir_uniform_ui(c, 24)); struct qreg src1_lo = qir_AND(c, src[1 * 4 + i], - qir_uniform_ui(c, 0xffff)); + qir_uniform_ui(c, 0xffffff)); struct qreg hilo = qir_MUL24(c, src0_hi, src1_lo); struct qreg lohi = qir_MUL24(c, src0_lo, src1_hi); @@ -381,7 +381,7 @@ tgsi_to_qir_umul(struct vc4_compile *c, return qir_ADD(c, lolo, qir_SHL(c, qir_ADD(c, hilo, lohi), - qir_uniform_ui(c, 16))); + qir_uniform_ui(c, 24))); } static struct qreg |