[Bug gprofng/32892] source line level information not available with "-g -O2"

2025-05-08 Thread sam at gentoo dot org 
https://sourceware.org/bugzilla/show_bug.cgi?id=32892

Sam James  changed:

   What|Removed |Added

   Target Milestone|--- |2.45

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gprofng/32892] source line level information not available with "-g -O2"

2025-05-08 Thread vladimir.mezentsev at oracle dot com 
https://sourceware.org/bugzilla/show_bug.cgi?id=32892

Vladimir Mezentsev  changed:

   What|Removed |Added

 Resolution|--- |FIXED
 Status|ASSIGNED|RESOLVED

--- Comment #2 from Vladimir Mezentsev  
---
Update status as resolved/fixed.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug libctf/32903] Error pointer overwritten on successful dict open in ctf_dict_open

2025-05-08 Thread nick.alcock at oracle dot com 
https://sourceware.org/bugzilla/show_bug.cgi?id=32903

Nick Alcock  changed:

   What|Removed |Added

   Assignee|unassigned at sourceware dot org   |nick.alcock at oracle 
dot com
 Ever confirmed|0   |1
 Status|UNCONFIRMED |ASSIGNED
   Last reconfirmed||2025-05-08

--- Comment #3 from Nick Alcock  ---
(Note that this does not trigger valgrind errors because we do not *do*
anything with the resulting err: in particular, it is never used to guide any
conditional branching. Callers that -- incorrectly, but understandably -- use
the err value despite there being no error would get hit with the valgrind
errors we escaped here.)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gprofng/32892] source line level information not available with "-g -O2"

2025-05-08 Thread vladimir.mezentsev at oracle dot com 
https://sourceware.org/bugzilla/show_bug.cgi?id=32892

--- Comment #4 from Vladimir Mezentsev  
---
 I cannot reproduce the problem:
% history
 1004  tar xf gperf-3.3.tar.gz 
 1005  cd gperf-3.3/
 1006  export CC=gcc; export CXX=g++; unset CFLAGS; unset CXXFLAGS; export
CPPFLAGS=-Wall
 1007  g++ --version
 1008  ./configure
 1009  make
 1010  cd src
 1011  gp-collect-app -o test.1.er -p on -S on ./gperf -m 10
../tests/5000.gperf > /dev/null

I have g++ 13.3.1:
% g++ --version
g++ (GCC) 13.3.1 20240611 (Red Hat 13.3.1-2.2.0.1)
Copyright (C) 2023 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

% gp-display-text -lines test.1.er/ | grep find_asso_values
 9.707  32.02   9.707  32.02   Search::find_asso_values(), line 47 in alternate
source context "bool-array.icc"
 9.567  31.56   9.567  31.56   Search::find_asso_values(), line 1400 in
"search.cc"
 2.812   9.28   2.812   9.28   Search::find_asso_values(), line 1399 in
"search.cc"
 2.782   9.18   2.782   9.18   Search::find_asso_values(), line 1396 in
"search.cc"
 1.031   3.40   1.031   3.40   Search::find_asso_values(), line 121 in
alternate source context "arraylist.h"
 0.951   3.14   0.951   3.14   Search::find_asso_values(), line 1401 in
"search.cc"
 0.690   2.28   0.690   2.28   Search::find_asso_values(), line 1397 in
"search.cc"
 0.170   0.56   0.170   0.56   Search::find_asso_values(), line 1407 in
"search.cc"
 0.100   0.33   0.100   0.33   Search::find_asso_values(), line 1390 in
"search.cc"
 0.050   0.17   0.050   0.17   Search::find_asso_values(), line 1398 in
"search.cc"
 0.040   0.13   0.040   0.13   Search::find_asso_values(), line 1475 in
"search.cc"
 0.030   0.10   0.030   0.10   Search::find_asso_values(), line 1523 in
"search.cc"
 0.020   0.07   0.020   0.07   Search::find_asso_values(), line 1473 in
"search.cc"
 0.020   0.07   0.020   0.07   Search::find_asso_values(), line 76 in alternate
source context "bool-array.icc"
 0.010   0.03   0.010   0.03   Search::find_asso_values(), line 1373 in
"search.cc"
 0.010   0.03   0.010   0.03   Search::find_asso_values(), line 1388 in
"search.cc"
 0.010   0.03   0.010   0.03   Search::find_asso_values(), line 1420 in
"search.cc"
 0.010   0.03   0.010   0.03   Search::find_asso_values(), line 1496 in
"search.cc"
 0.010   0.03   0.010   0.03   Search::find_asso_values(), line 1507 in
"search.cc"
 0.010   0.03   0.010   0.03   Search::find_asso_values(), line 1509 in
"search.cc"
 0.010   0.03   0.010   0.03   Search::find_asso_values(), line 66 in alternate
source context "bool-array.icc"
 0.  0. 0.030   0.10   Search::find_asso_values(), line 1099 in
"search.cc"
 0.  0. 0.020   0.07   Search::find_asso_values(), line 1101 in
"search.cc"
 0.  0. 0.500   1.65   Search::find_asso_values(), line 1197 in
"search.cc"
 0.  0. 1.101   3.63   Search::find_asso_values(), line 1217 in
"search.cc"
 0.  0. 0.310   1.02   Search::find_asso_values(), line 1245 in
"search.cc"
 0.  0. 0.050   0.17   Search::find_asso_values(), line 1255 in
"search.cc"
 0.  0. 0.040   0.13   Search::find_asso_values(), line 69 in alternate
source context "bool-array.icc"



If your $PATH was correct and you used the latest version of gprofng, could you
attach or send me directly your ./gperf

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gprofng/32892] source line level information not available with "-g -O2"

2025-05-08 Thread bruno at clisp dot org 
https://sourceware.org/bugzilla/show_bug.cgi?id=32892

Bruno Haible  changed:

   What|Removed |Added

 Resolution|FIXED   |---
 Status|RESOLVED|REOPENED

--- Comment #3 from Bruno Haible  ---
Sorry to say so, but I see no improvement, compared to binutils 2.44.

I build the current binutils-gdb tree (from today) from source, installed it,
and am using it via PATH.

The result is still:

8. gp-collect-app -o test.1.er -p on -S on ./gperf -m 10 ../tests/5000.gperf >
/dev/null
9. gprofng-display-text test.1.er
(gprofng-display-text) lines
...
13.479  92.07  14.440  98.63   
...
(gprofng-display-text) source search.cc

Expected:
Lines search.cc 1396..1402 are highlighted as CPU-intensive.

Actual:
Lines search.cc 1396..1402 are not marked in any way.

This is on Linux/x86_64, with g++ 13.3.0 and like with g++ 15.1.0.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug ld/17112] [regression] binutils trunk ld/LTO 11 test fails when built using GCC 4.9

2025-05-08 Thread sam at gentoo dot org 
https://sourceware.org/bugzilla/show_bug.cgi?id=17112

Sam James  changed:

   What|Removed |Added

   See Also||https://gcc.gnu.org/bugzill
   ||a/show_bug.cgi?id=61715

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gprofng/32892] source line level information not available with "-g -O2"

2025-05-08 Thread bruno at clisp dot org 
https://sourceware.org/bugzilla/show_bug.cgi?id=32892

--- Comment #6 from Bruno Haible  ---
The binutils binaries that I built today identify themselves as

$ gp-collect-app --version
GNU gprofng-collect-app binutils version 2.44.50
Copyright (C) 2025 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later .
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
$ gp-display-text --version
GNU gprofng-display-text binutils version 2.44.50
Copyright (C) 2025 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later .
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gprofng/32892] source line level information not available with "-g -O2"

2025-05-08 Thread bruno at clisp dot org 
https://sourceware.org/bugzilla/show_bug.cgi?id=32892

--- Comment #5 from Bruno Haible  ---
Created attachment 16076
  --> https://sourceware.org/bugzilla/attachment.cgi?id=16076&action=edit
gperf and test.1.er from Bruno's machine

I got the same negative results with various GCC versions:

g++ (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0
Copyright (C) 2023 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

g++ (GCC) 13.3.0
Copyright (C) 2023 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

g++ (GCC) 14.2.0
Copyright (C) 2024 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

g++ (GCC) 15.1.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Find attached here (as brunos-binaries.tar.gz) the 'gperf' program (created
with the 'Ubuntu 13.3.0-6ubuntu2~24.04' compiler) and the test.1.er directory
from the gperf-3.3/src/ directory.

1) If you run
$ gp-display-text -lines test.1.er/ | grep find_asso_values
and it displays source line numbers, the difference between your and my machine
is in the gp-display-text step.

2) If you run
$ rm -rf test.1.er/
$ gp-collect-app -o test.1.er -p on -S on ./gperf -m 10 ../tests/5000.gperf >
/dev/null
$ gp-display-text -lines test.1.er/ | grep find_asso_values
and it displays source line numbers, the difference between your and my machine
is in the gp-collect-app step.

3) Otherwise, the difference between your and my machine is in the 'gperf'
executable. Maybe yours contains different types of debug information?

-- 
You are receiving this mail because:
You are on the CC list for the bug.