lkml.org 
[lkml]   [2019]   [Jun]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH 08/11] perf intel-pt: Add xmm registers to synthesized PEBS sample
Date
Add xmm register information from PEBS data in the Intel PT trace to the
synthesized PEBS sample.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
tools/perf/util/intel-pt.c | 30 +++++++++++++++++++++++++++++-
1 file changed, 29 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 3a6d5af6a71b..0175a4ec248f 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1499,6 +1499,31 @@ static u64 *intel_pt_add_gp_regs(struct regs_dump *intr_regs, u64 *pos,
return pos;
}

+#ifndef PERF_REG_X86_XMM0
+#define PERF_REG_X86_XMM0 32
+#endif
+
+static void intel_pt_add_xmm(struct regs_dump *intr_regs, u64 *pos,
+ const struct intel_pt_blk_items *items,
+ u64 regs_mask)
+{
+ u32 mask = items->has_xmm & (regs_mask >> PERF_REG_X86_XMM0);
+ const u64 *xmm = items->xmm;
+
+ /*
+ * If there are any XMM registers, then there should be all of them.
+ * Nevertheless, follow the logic to add only registers that were
+ * requested (i.e. 'regs_mask') and that were provided (i.e. 'mask'),
+ * and update the resulting mask (i.e. 'intr_regs->mask') accordingly.
+ */
+ intr_regs->mask |= (u64)mask << PERF_REG_X86_XMM0;
+
+ for (; mask; mask >>= 1, xmm++) {
+ if (mask & 1)
+ *pos++ = *xmm;
+ }
+}
+
static int intel_pt_synth_pebs_sample(struct intel_pt_queue *ptq)
{
const struct intel_pt_blk_items *items = &ptq->state->items;
@@ -1553,13 +1578,16 @@ static int intel_pt_synth_pebs_sample(struct intel_pt_queue *ptq)
items->mask[INTEL_PT_GP_REGS_POS]) {
u64 regs[sizeof(sample.intr_regs.mask)];
u64 regs_mask = evsel->attr.sample_regs_intr;
+ u64 *pos;

sample.intr_regs.abi = items->is_32_bit ?
PERF_SAMPLE_REGS_ABI_32 :
PERF_SAMPLE_REGS_ABI_64;
sample.intr_regs.regs = regs;

- intel_pt_add_gp_regs(&sample.intr_regs, regs, items, regs_mask);
+ pos = intel_pt_add_gp_regs(&sample.intr_regs, regs, items, regs_mask);
+
+ intel_pt_add_xmm(&sample.intr_regs, pos, items, regs_mask);
}

return intel_pt_deliver_synth_event(pt, ptq, event, &sample, sample_type);
--
2.17.1
\
 
 \ /
  Last update: 2019-06-10 09:30    [W:0.129 / U:0.028 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site