public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Hannes Domani <ssbssa@yahoo.de>
To: gdb-patches@sourceware.org
Subject: [PATCH] Fix backtrace limit stopping on inline frame
Date: Sat,  6 Jan 2024 15:17:35 +0100	[thread overview]
Message-ID: <20240106141735.2330-1-ssbssa@yahoo.de> (raw)
In-Reply-To: <20240106141735.2330-1-ssbssa.ref@yahoo.de>

If you have set up a backtrace limit, and the backtrace stops
because of this in an inline frame with arguments, you get an
assertion failure:
```
(gdb) bt
#0  normal_frame (i=0) at gdb-29865.c:4
#1  0x000000013fe3162a in inline_frame (i=0) at gdb-29865.c:9
#2  main () at gdb-29865.c:14
(gdb) set backtrace limit 2
(gdb) bt
#0  normal_frame (i=0) at gdb-29865.c:4
#1  0x000000013fe3162a in inline_frame (
C:/src/repos/binutils-gdb.git/gdb/frame.c:3346: internal-error: reinflate: Assertion `m_cached_level >= -1' failed.
```

And if this one if fixed, there is another one as well:
```
(gdb) bt
#0  normal_frame (i=0) at gdb-29865.c:4
#1  0x000000013fdf162a in inline_frame (
C:/src/repos/binutils-gdb.git/gdb/dwarf2/loc.c:1160: internal-error: dwarf_expr_reg_to_entry_parameter: Assertion `frame != NULL' failed.
```

The reason for both of them is this kind of loop:
```
  while (get_frame_type (frame) == INLINE_FRAME)
    frame = get_prev_frame (frame);
```
Since get_prev_frame respects the backtrace limit, it will return
NULL, and from there on you can't continue.
This changes these loops to use get_prev_frame_always instead, so
you always get a non-inline frame in the end.

With this backtrace works:
```
(gdb) bt
#0  normal_frame (i=0) at gdb-29865.c:4
#1  0x000000013fd4162a in inline_frame (i=0) at gdb-29865.c:9
(gdb)
```

Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=29865
---
 gdb/dwarf2/frame.c                  | 2 +-
 gdb/dwarf2/loc.c                    | 2 +-
 gdb/testsuite/gdb.opt/inline-bt.c   | 8 ++++----
 gdb/testsuite/gdb.opt/inline-bt.exp | 1 +
 4 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/gdb/dwarf2/frame.c b/gdb/dwarf2/frame.c
index d3d1ecdf1f5..143b934e5ef 100644
--- a/gdb/dwarf2/frame.c
+++ b/gdb/dwarf2/frame.c
@@ -1423,7 +1423,7 @@ dwarf2_frame_cfa (frame_info_ptr this_frame)
 		 _("cfa not available for record btrace target"));
 
   while (get_frame_type (this_frame) == INLINE_FRAME)
-    this_frame = get_prev_frame (this_frame);
+    this_frame = get_prev_frame_always (this_frame);
   if (get_frame_unwind_stop_reason (this_frame) == UNWIND_UNAVAILABLE)
     throw_error (NOT_AVAILABLE_ERROR,
 		_("can't compute CFA for this frame: "
diff --git a/gdb/dwarf2/loc.c b/gdb/dwarf2/loc.c
index c15221eb7a2..b1f188bf9c5 100644
--- a/gdb/dwarf2/loc.c
+++ b/gdb/dwarf2/loc.c
@@ -1156,7 +1156,7 @@ dwarf_expr_reg_to_entry_parameter (frame_info_ptr frame,
 
   while (get_frame_type (frame) == INLINE_FRAME)
     {
-      frame = get_prev_frame (frame);
+      frame = get_prev_frame_always (frame);
       gdb_assert (frame != NULL);
     }
 
diff --git a/gdb/testsuite/gdb.opt/inline-bt.c b/gdb/testsuite/gdb.opt/inline-bt.c
index 8dac8d30300..0dad0f47e6e 100644
--- a/gdb/testsuite/gdb.opt/inline-bt.c
+++ b/gdb/testsuite/gdb.opt/inline-bt.c
@@ -28,15 +28,15 @@ volatile int result;
 
 void bar(void);
 
-inline ATTR int func1(void)
+inline ATTR int func1(int s)
 {
   bar ();
-  return x * y;
+  return x * y + s;
 }
 
 inline ATTR int func2(void)
 {
-  return x * func1 ();
+  return x * func1 (1);
 }
 
 int main (void)
@@ -47,7 +47,7 @@ int main (void)
   y = 8;
   bar ();
 
-  val = func1 ();
+  val = func1 (2);
   result = val;
 
   val = func2 ();
diff --git a/gdb/testsuite/gdb.opt/inline-bt.exp b/gdb/testsuite/gdb.opt/inline-bt.exp
index 501b24109e3..b0e5f61a5c7 100644
--- a/gdb/testsuite/gdb.opt/inline-bt.exp
+++ b/gdb/testsuite/gdb.opt/inline-bt.exp
@@ -65,3 +65,4 @@ gdb_test "up" "#1  .*func1.*" "up from bar (4)"
 gdb_test "info frame" ".*in func1.*" "info frame still works"
 # Verify the user visible limit works as expected.
 gdb_test "up" "Initial frame selected; you cannot go up." "up hits limit"
+gdb_test "backtrace" "#0  bar.*#1  .*func1.*" "backtrace hits limit"
-- 
2.35.1


       reply	other threads:[~2024-01-06 14:18 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20240106141735.2330-1-ssbssa.ref@yahoo.de>
2024-01-06 14:17 ` Hannes Domani [this message]
2024-01-17 16:05   ` [PING] " Hannes Domani
2024-01-26 17:05     ` Hannes Domani
2024-01-29 10:55   ` Andrew Burgess
2024-01-29 14:35     ` Hannes Domani

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240106141735.2330-1-ssbssa@yahoo.de \
    --to=ssbssa@yahoo.de \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).