[Mesa-dev] [PATCH 5/7] nvc0: remove useless counting operations for MP counters
Samuel Pitoiset
samuel.pitoiset at gmail.com
Wed Dec 2 14:49:49 PST 2015
Those bits were related to old performance metrics support.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset at gmail.com>
---
.../drivers/nouveau/nvc0/nvc0_query_hw_sm.c | 106 +--------------------
1 file changed, 5 insertions(+), 101 deletions(-)
diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw_sm.c b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw_sm.c
index feec431..7d1e75f 100644
--- a/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw_sm.c
+++ b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw_sm.c
@@ -185,24 +185,15 @@ struct nvc0_hw_sm_counter_cfg
uint32_t src_sel; /* signal selection for up to 4 sources */
};
-#define NVC0_COUNTER_OPn_SUM 0
-#define NVC0_COUNTER_OPn_OR 1
-#define NVC0_COUNTER_OPn_AND 2
-#define NVC0_COUNTER_OP2_REL_SUM_MM 3 /* (sum(ctr0) - sum(ctr1)) / sum(ctr0) */
-#define NVC0_COUNTER_OP2_DIV_SUM_M0 4 /* sum(ctr0) / ctr1 of MP[0]) */
-#define NVC0_COUNTER_OP2_AVG_DIV_MM 5 /* avg(ctr0 / ctr1) */
-#define NVC0_COUNTER_OP2_AVG_DIV_M0 6 /* avg(ctr0) / ctr1 of MP[0]) */
-
struct nvc0_hw_sm_query_cfg
{
struct nvc0_hw_sm_counter_cfg ctr[8];
uint8_t num_counters;
- uint8_t op;
uint8_t norm[2]; /* normalization num,denom */
};
-#define _Q1A(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g, 0, s }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { nu, dn } }
-#define _Q1B(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g, 0, s }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { nu, dn } }
+#define _Q1A(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g, 0, s }, {}, {}, {} }, 1, { nu, dn } }
+#define _Q1B(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g, 0, s }, {}, {}, {} }, 1, { nu, dn } }
/* NOTES:
* active_warps: bit 0 alternates btw 0 and 1 for odd nr of warps
@@ -362,7 +353,6 @@ sm20_active_cycles =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x11, 0x000000ff, 0x00000000),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -376,7 +366,6 @@ sm20_active_warps =
.ctr[4] = _C(0xaaaa, LOGOP, 0x24, 0x000000ff, 0x00000050),
.ctr[5] = _C(0xaaaa, LOGOP, 0x24, 0x000000ff, 0x00000060),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -385,7 +374,6 @@ sm20_atom_count =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x63, 0x000000ff, 0x00000030),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -395,7 +383,6 @@ sm20_branch =
.ctr[0] = _C(0xaaaa, LOGOP, 0x1a, 0x000000ff, 0x00000000),
.ctr[1] = _C(0xaaaa, LOGOP, 0x1a, 0x000000ff, 0x00000010),
.num_counters = 2,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -405,7 +392,6 @@ sm20_divergent_branch =
.ctr[0] = _C(0xaaaa, LOGOP, 0x19, 0x000000ff, 0x00000020),
.ctr[1] = _C(0xaaaa, LOGOP, 0x19, 0x000000ff, 0x00000030),
.num_counters = 2,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -414,7 +400,6 @@ sm20_gld_request =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x64, 0x000000ff, 0x00000030),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -423,7 +408,6 @@ sm20_gred_count =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x63, 0x000000ff, 0x00000040),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -432,7 +416,6 @@ sm20_gst_request =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x64, 0x000000ff, 0x00000060),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -442,7 +425,6 @@ sm20_inst_executed =
.ctr[0] = _C(0xaaaa, LOGOP, 0x2d, 0x0000ffff, 0x00001000),
.ctr[1] = _C(0xaaaa, LOGOP, 0x2d, 0x0000ffff, 0x00001010),
.num_counters = 2,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -452,7 +434,6 @@ sm20_inst_issued =
.ctr[0] = _C(0xaaaa, LOGOP, 0x27, 0x0000ffff, 0x00007060),
.ctr[1] = _C(0xaaaa, LOGOP, 0x27, 0x0000ffff, 0x00007070),
.num_counters = 2,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -461,7 +442,6 @@ sm20_local_ld =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x64, 0x000000ff, 0x00000020),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -470,7 +450,6 @@ sm20_local_st =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x64, 0x000000ff, 0x00000050),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -479,7 +458,6 @@ sm20_prof_trigger_0 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000000),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -488,7 +466,6 @@ sm20_prof_trigger_1 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000010),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -497,7 +474,6 @@ sm20_prof_trigger_2 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000020),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -506,7 +482,6 @@ sm20_prof_trigger_3 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000030),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -515,7 +490,6 @@ sm20_prof_trigger_4 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000040),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -524,7 +498,6 @@ sm20_prof_trigger_5 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000050),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -533,7 +506,6 @@ sm20_prof_trigger_6 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000060),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -542,7 +514,6 @@ sm20_prof_trigger_7 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x01, 0x000000ff, 0x00000070),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -551,7 +522,6 @@ sm20_shared_ld =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x64, 0x000000ff, 0x00000010),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -560,7 +530,6 @@ sm20_shared_st =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x64, 0x000000ff, 0x00000040),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -574,7 +543,6 @@ sm20_threads_launched =
.ctr[4] = _C(0xaaaa, LOGOP, 0x26, 0x000000ff, 0x00000050),
.ctr[5] = _C(0xaaaa, LOGOP, 0x26, 0x000000ff, 0x00000060),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -588,7 +556,6 @@ sm20_th_inst_executed_0 =
.ctr[4] = _C(0xaaaa, LOGOP, 0x2f, 0x000000ff, 0x00000040),
.ctr[5] = _C(0xaaaa, LOGOP, 0x2f, 0x000000ff, 0x00000050),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -602,7 +569,6 @@ sm20_th_inst_executed_1 =
.ctr[4] = _C(0xaaaa, LOGOP, 0x30, 0x000000ff, 0x00000040),
.ctr[5] = _C(0xaaaa, LOGOP, 0x30, 0x000000ff, 0x00000050),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -611,7 +577,6 @@ sm20_warps_launched =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x26, 0x000000ff, 0x00000000),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -659,7 +624,6 @@ sm21_inst_executed =
.ctr[1] = _C(0xaaaa, LOGOP, 0x2d, 0x000000ff, 0x00000010),
.ctr[2] = _C(0xaaaa, LOGOP, 0x2d, 0x000000ff, 0x00000020),
.num_counters = 3,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -668,7 +632,6 @@ sm21_inst_issued1_0 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x7e, 0x000000ff, 0x00000010),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -677,7 +640,6 @@ sm21_inst_issued1_1 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x7e, 0x000000ff, 0x00000040),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -686,7 +648,6 @@ sm21_inst_issued2_0 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x7e, 0x000000ff, 0x00000020),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -695,7 +656,6 @@ sm21_inst_issued2_1 =
{
.ctr[0] = _C(0xaaaa, LOGOP, 0x7e, 0x000000ff, 0x00000050),
.num_counters = 1,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -709,7 +669,6 @@ sm21_th_inst_executed_0 =
.ctr[4] = _C(0xaaaa, LOGOP, 0xa3, 0x000000ff, 0x00000040),
.ctr[5] = _C(0xaaaa, LOGOP, 0xa3, 0x000000ff, 0x00000050),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -723,7 +682,6 @@ sm21_th_inst_executed_1 =
.ctr[4] = _C(0xaaaa, LOGOP, 0xa5, 0x000000ff, 0x00000040),
.ctr[5] = _C(0xaaaa, LOGOP, 0xa5, 0x000000ff, 0x00000050),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -737,7 +695,6 @@ sm21_th_inst_executed_2 =
.ctr[4] = _C(0xaaaa, LOGOP, 0xa4, 0x000000ff, 0x00000040),
.ctr[5] = _C(0xaaaa, LOGOP, 0xa4, 0x000000ff, 0x00000050),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -751,7 +708,6 @@ sm21_th_inst_executed_3 =
.ctr[4] = _C(0xaaaa, LOGOP, 0xa6, 0x000000ff, 0x00000040),
.ctr[5] = _C(0xaaaa, LOGOP, 0xa6, 0x000000ff, 0x00000050),
.num_counters = 6,
- .op = NVC0_COUNTER_OPn_SUM,
.norm = { 1, 1 },
};
@@ -1150,62 +1106,10 @@ nvc0_hw_sm_get_query_result(struct nvc0_context *nvc0, struct nvc0_hw_query *hq,
if (!ret)
return false;
- if (cfg->op == NVC0_COUNTER_OPn_SUM) {
- for (c = 0; c < cfg->num_counters; ++c)
- for (p = 0; p < mp_count; ++p)
- value += count[p][c];
- value = (value * cfg->norm[0]) / cfg->norm[1];
- } else
- if (cfg->op == NVC0_COUNTER_OPn_OR) {
- uint32_t v = 0;
- for (c = 0; c < cfg->num_counters; ++c)
- for (p = 0; p < mp_count; ++p)
- v |= count[p][c];
- value = ((uint64_t)v * cfg->norm[0]) / cfg->norm[1];
- } else
- if (cfg->op == NVC0_COUNTER_OPn_AND) {
- uint32_t v = ~0;
- for (c = 0; c < cfg->num_counters; ++c)
- for (p = 0; p < mp_count; ++p)
- v &= count[p][c];
- value = ((uint64_t)v * cfg->norm[0]) / cfg->norm[1];
- } else
- if (cfg->op == NVC0_COUNTER_OP2_REL_SUM_MM) {
- uint64_t v[2] = { 0, 0 };
- for (p = 0; p < mp_count; ++p) {
- v[0] += count[p][0];
- v[1] += count[p][1];
- }
- if (v[0])
- value = ((v[0] - v[1]) * cfg->norm[0]) / (v[0] * cfg->norm[1]);
- } else
- if (cfg->op == NVC0_COUNTER_OP2_DIV_SUM_M0) {
+ for (c = 0; c < cfg->num_counters; ++c)
for (p = 0; p < mp_count; ++p)
- value += count[p][0];
- if (count[0][1])
- value = (value * cfg->norm[0]) / (count[0][1] * cfg->norm[1]);
- else
- value = 0;
- } else
- if (cfg->op == NVC0_COUNTER_OP2_AVG_DIV_MM) {
- unsigned mp_used = 0;
- for (p = 0; p < mp_count; ++p, mp_used += !!count[p][0])
- if (count[p][1])
- value += (count[p][0] * cfg->norm[0]) / count[p][1];
- if (mp_used)
- value /= (uint64_t)mp_used * cfg->norm[1];
- } else
- if (cfg->op == NVC0_COUNTER_OP2_AVG_DIV_M0) {
- unsigned mp_used = 0;
- for (p = 0; p < mp_count; ++p, mp_used += !!count[p][0])
- value += count[p][0];
- if (count[0][1] && mp_used) {
- value *= cfg->norm[0];
- value /= (uint64_t)count[0][1] * mp_used * cfg->norm[1];
- } else {
- value = 0;
- }
- }
+ value += count[p][c];
+ value = (value * cfg->norm[0]) / cfg->norm[1];
*(uint64_t *)result = value;
return true;
--
2.6.2
More information about the mesa-dev
mailing list