twx-linux/tools/perf
Namhyung Kim 1cfd01eb60 perf annotate-data: Copy back variable types after move
In some cases, compilers don't set the location expression in DWARF
precisely.  For instance, it may assign a variable to a register after
copying it from a different register.  Then it should use the register
for the new type but still uses the old register.  This makes hard to
track the type information properly.

This is an example I found in __tcp_transmit_skb().  The first argument
(sk) of this function is a pointer to sock and there's a variable (tp)
for tcp_sock.

  static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb,
  				int clone_it, gfp_t gfp_mask, u32 rcv_nxt)
  {
  	...
  	struct tcp_sock *tp;

  	BUG_ON(!skb || !tcp_skb_pcount(skb));
  	tp = tcp_sk(sk);
  	prior_wstamp = tp->tcp_wstamp_ns;
  	tp->tcp_wstamp_ns = max(tp->tcp_wstamp_ns, tp->tcp_clock_cache);
  	...

So it basically calls tcp_sk(sk) to get the tcp_sock pointer from sk.
But it turned out to be the same value because tcp_sock embeds sock as
the first member.  The sk is located in reg5 (RDI) and tp is in reg3
(RBX).  The offset of tcp_wstamp_ns is 0x748 and tcp_clock_cache is
0x750.  So you need to use RBX (reg3) to access the fields in the
tcp_sock.  But the code used RDI (reg5) as it has the same value.

  $ pahole --hex -C tcp_sock vmlinux | grep -e 748 -e 750
	u64                tcp_wstamp_ns;        /* 0x748   0x8 */
	u64                tcp_clock_cache;      /* 0x750   0x8 */

And this is the disassembly of the part of the function.

  <__tcp_transmit_skb>:
  ...
  44:  mov    %rdi, %rbx
  47:  mov    0x748(%rdi), %rsi
  4e:  mov    0x750(%rdi), %rax
  55:  cmp    %rax, %rsi

Because compiler put the debug info to RBX, it only knows RDI is a
pointer to sock and accessing those two fields resulted in error
due to offset being beyond the type size.

  -----------------------------------------------------------
  find data type for 0x748(reg5) at __tcp_transmit_skb+0x63
  CU for net/ipv4/tcp_output.c (die:0x817f543)
  frame base: cfa=0 fbreg=6
  scope: [1/1] (die:81aac3e)
  bb: [0 - 30]
  var [0] -0x98(stack) type='struct tcp_out_options' size=0x28 (die:0x81af3df)
  var [5] reg8 type='unsigned int' size=0x4 (die:0x8180ed6)
  var [5] reg2 type='unsigned int' size=0x4 (die:0x8180ed6)
  var [5] reg1 type='int' size=0x4 (die:0x818059e)
  var [5] reg4 type='struct sk_buff*' size=0x8 (die:0x8181360)
  var [5] reg5 type='struct sock*' size=0x8 (die:0x8181a0c)                   <<<--- the first argument ('sk' at %RDI)
  mov [19] reg8 -> -0xa8(stack) type='unsigned int' size=0x4 (die:0x8180ed6)
  mov [20] stack canary -> reg0
  mov [29] reg0 -> -0x30(stack) stack canary
  bb: [36 - 3e]
  mov [36] reg4 -> reg15 type='struct sk_buff*' size=0x8 (die:0x8181360)
  bb: [44 - 63]
  mov [44] reg5 -> reg3 type='struct sock*' size=0x8 (die:0x8181a0c)          <<<--- calling tcp_sk()
  var [47] reg3 type='struct tcp_sock*' size=0x8 (die:0x819eead)              <<<--- new variable ('tp' at %RBX)
  var [4e] reg4 type='unsigned long long' size=0x8 (die:0x8180edd)
  mov [58] reg4 -> -0xc0(stack) type='unsigned long long' size=0x8 (die:0x8180edd)
  chk [63] reg5 offset=0x748 ok=1 kind=1 (struct sock*) : offset bigger than size    <<<--- access with old variable
  final result: offset bigger than size

While it's a fault in the compiler, we could work around this issue by
using the type of new variable when it's copied directly.  So I've added
copied_from field in the register state to track those direct register
to register copies.  After that new register gets a new type and the old
register still has the same type, it'll update (copy it back) the type
of the old register.

For example, if we can update type of reg5 at __tcp_transmit_skb+0x47,
we can find the target type of the instruction at 0x63 like below:

  -----------------------------------------------------------
  find data type for 0x748(reg5) at __tcp_transmit_skb+0x63
  ...
  bb: [44 - 63]
  mov [44] reg5 -> reg3 type='struct sock*' size=0x8 (die:0x8181a0c)
  var [47] reg3 type='struct tcp_sock*' size=0x8 (die:0x819eead)
  var [47] copyback reg5 type='struct tcp_sock*' size=0x8 (die:0x819eead)     <<<--- here
  mov [47] 0x748(reg5) -> reg4 type='unsigned long long' size=0x8 (die:0x8180edd)
  mov [4e] 0x750(reg5) -> reg0 type='unsigned long long' size=0x8 (die:0x8180edd)
  mov [58] reg4 -> -0xc0(stack) type='unsigned long long' size=0x8 (die:0x8180edd)
  chk [63] reg5 offset=0x748 ok=1 kind=1 (struct tcp_sock*) : Good!           <<<--- new type
  found by insn track: 0x748(reg5) type-offset=0x748
  final result:  type='struct tcp_sock' size=0xa98 (die:0x819eeb2)

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20240821232628.353177-5-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2024-08-22 12:38:18 -03:00
..
arch perf annotate-data: Copy back variable types after move 2024-08-22 12:38:18 -03:00
bench perf tool: Constify tool pointers 2024-08-12 18:05:14 -03:00
dlfilters
Documentation perf script: Add branch counters 2024-08-14 10:20:40 -03:00
include/perf
jvmti
pmu-events perf vendor events intel: Add MTL metric JSON files 2024-08-13 15:25:32 -03:00
python
scripts perf scripts python cs-etm: Restore first sample log in verbose mode 2024-07-31 16:58:18 -03:00
tests perf test: Add cgroup sampling test 2024-08-19 16:32:32 -03:00
trace Merge remote-tracking branch 'torvalds/master' into perf-tools-next 2024-08-16 19:43:16 -03:00
ui perf annotate-data: Show offset and size in hex 2024-08-21 11:48:39 -03:00
util perf annotate-data: Copy back variable types after move 2024-08-22 12:38:18 -03:00
.gitignore
Build perf util: Make util its own library 2024-06-26 11:07:42 -07:00
builtin-annotate.c perf annotate: Display the branch counter histogram 2024-08-14 10:20:40 -03:00
builtin-bench.c perf bench uprobe: Add uretprobe variant of uprobe benchmarks 2024-04-12 17:54:02 -03:00
builtin-buildid-cache.c perf dso: Add reference count checking and accessor functions 2024-05-06 15:28:49 -03:00
builtin-buildid-list.c perf buildid-list: Use perf_tool__init 2024-08-12 18:07:10 -03:00
builtin-c2c.c perf c2c: Use perf_tool__init() 2024-08-12 18:09:32 -03:00
builtin-config.c
builtin-daemon.c perf daemon: Fix the build on more 32-bit architectures 2024-08-19 21:44:30 -03:00
builtin-data.c
builtin-diff.c perf report: Display the branch counter histogram 2024-08-14 10:20:40 -03:00
builtin-evlist.c perf evlist: Use perf_tool__init() 2024-08-12 18:08:35 -03:00
builtin-ftrace.c perf cap: Tidy up and improve capability testing 2024-08-20 17:53:12 -03:00
builtin-help.c
builtin-inject.c perf inject: Combine mmap and mmap2 handling 2024-08-19 14:57:15 -03:00
builtin-kallsyms.c perf dso: Add reference count checking and accessor functions 2024-05-06 15:28:49 -03:00
builtin-kmem.c perf kmem: Use perf_tool__init 2024-08-12 18:06:48 -03:00
builtin-kvm.c perf kvm: Use perf_tool__init() 2024-08-12 18:07:40 -03:00
builtin-kwork.c perf kwork: Use perf_tool__init() 2024-08-12 18:13:39 -03:00
builtin-list.c perf list: Fix the --no-desc option 2024-05-28 11:29:49 -07:00
builtin-lock.c perf lock: Use perf_tool__init() 2024-08-12 18:08:08 -03:00
builtin-mem.c perf mem: Use perf_tool__init() 2024-08-12 18:11:49 -03:00
builtin-probe.c perf probe: Use zfree() to avoid possibly accessing dangling pointers 2024-05-07 12:44:02 -03:00
builtin-record.c perf record: Fix sample cgroup & namespace tracking 2024-08-19 16:32:05 -03:00
builtin-report.c perf report: Display the branch counter histogram 2024-08-14 10:20:40 -03:00
builtin-sched.c perf sched timehist: Fix missing free of session in perf_sched__timehist() 2024-08-16 19:31:15 -03:00
builtin-script.c perf script: Add branch counters 2024-08-14 10:20:40 -03:00
builtin-stat.c perf stat: Add command line option for enabling TPEBS recording 2024-08-13 15:25:32 -03:00
builtin-timechart.c perf timechart: Use perf_tool__init() 2024-08-12 18:12:06 -03:00
builtin-top.c perf map: API clean up 2024-08-19 14:49:53 -03:00
builtin-trace.c perf tool: Constify tool pointers 2024-08-12 18:05:14 -03:00
builtin-version.c
builtin.h
check-headers.sh tools headers: Synchronize linux/bits.h with the kernel sources 2024-04-26 22:13:10 -03:00
command-list.txt
CREDITS
design.txt
Makefile perf tools: Fix wrong message when running "make JOBS=1" 2024-08-01 12:11:33 -03:00
Makefile.config perf cap: Tidy up and improve capability testing 2024-08-20 17:53:12 -03:00
Makefile.perf perf tools: Create source symlink in perf object dir 2024-08-09 17:37:24 -03:00
MANIFEST
perf-archive.sh
perf-completion.sh
perf-iostat.sh
perf-read-vdso.c
perf-sys.h
perf.c
perf.h