perf intel-pt: Add support for emulated ptwrite

ptwrite is an Intel x86 instruction that writes arbitrary values into an
Intel PT trace. It is not supported on all hardware, so provide an
alternative that makes use of TNT packets to convey the payload data.
TNT packets encode Taken/Not-taken conditional branch information, so
taking branches based on the payload value will encode the value into
the TNT packet. Refer to the changes to the documentation file
perf-intel-pt.txt in this patch for an example.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/20220509152400.376613-2-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

1 files changed, 88 insertions, 0 deletions

diff --git a/tools/perf/Documentation/perf-intel-pt.txt b/tools/perf/Documentation/perf-intel-pt.txt
index 92532d0d3618..8acd704e8a39 100644
--- a/tools/perf/Documentation/perf-intel-pt.txt
+++ b/tools/perf/Documentation/perf-intel-pt.txt
@@ -468,6 +468,8 @@ ptw		Enable PTWRITE packets which are produced when a ptwrite instruction
 		which contains "1" if the feature is supported and
 		"0" otherwise.
 
+		As an alternative, refer to "Emulated PTWRITE" further below.
+
 fup_on_ptw	Enable a FUP packet to follow the PTWRITE packet.  The FUP packet
 		provides the address of the ptwrite instruction.  In the absence of
 		fup_on_ptw, the decoder will use the address of the previous branch
@@ -1471,6 +1473,92 @@ In that case the --itrace q option is forced because walking executable code
 to reconstruct the control flow is not possible.
 
 
+Emulated PTWRITE
+----------------
+
+Later perf tools support a method to emulate the ptwrite instruction, which
+can be useful if hardware does not support the ptwrite instruction.
+
+Instead of using the ptwrite instruction, a function is used which produces
+a trace that encodes the payload data into TNT packets.  Here is an example
+of the function:
+
+ #include <stdint.h>
+
+ void perf_emulate_ptwrite(uint64_t x)
+ __attribute__((externally_visible, noipa, no_instrument_function, naked));
+
+ #define PERF_EMULATE_PTWRITE_8_BITS \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"        \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"        \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"        \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"        \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"        \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"        \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"        \
+                 "1: shl %rax\n"     \
+                 "   jc 1f\n"
+
+ /* Undefined instruction */
+ #define PERF_EMULATE_PTWRITE_UD2        ".byte 0x0f, 0x0b\n"
+
+ #define PERF_EMULATE_PTWRITE_MAGIC        PERF_EMULATE_PTWRITE_UD2 ".ascii \"perf,ptwrite  \"\n"
+
+ void perf_emulate_ptwrite(uint64_t x __attribute__ ((__unused__)))
+ {
+          /* Assumes SysV ABI : x passed in rdi */
+         __asm__ volatile (
+                 "jmp 1f\n"
+                 PERF_EMULATE_PTWRITE_MAGIC
+                 "1: mov %rdi, %rax\n"
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 PERF_EMULATE_PTWRITE_8_BITS
+                 "1: ret\n"
+         );
+ }
+
+For example, a test program with the function above:
+
+ #include <stdio.h>
+ #include <stdint.h>
+ #include <stdlib.h>
+
+ #include "perf_emulate_ptwrite.h"
+
+ int main(int argc, char *argv[])
+ {
+         uint64_t x = 0;
+
+         if (argc > 1)
+                 x = strtoull(argv[1], NULL, 0);
+         perf_emulate_ptwrite(x);
+         return 0;
+ }
+
+Can be compiled and traced:
+
+ $ gcc -Wall -Wextra -O3 -g -o eg_ptw eg_ptw.c
+ $ perf record -e intel_pt//u ./eg_ptw 0x1234567890abcdef
+ [ perf record: Woken up 1 times to write data ]
+ [ perf record: Captured and wrote 0.017 MB perf.data ]
+ $ perf script --itrace=ew
+           eg_ptw 19875 [007]  8061.235912:     ptwrite:  IP: 0 payload: 0x1234567890abcdef      55701249a196 perf_emulate_ptwrite+0x16 (/home/user/eg_ptw)
+ $
+
+
 EXAMPLE
 -------