[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2024-05-03 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

Paul Floyd  changed:

   What|Removed |Added

 CC||limuyang202...@163.com

--- Comment #13 from Paul Floyd  ---
*** Bug 486538 has been marked as a duplicate of this bug. ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

Paul Floyd  changed:

   What|Removed |Added

 Status|REPORTED|RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Dennis Kobert
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #12 from Dennis Kobert  ---
(In reply to Paul Floyd from comment #11)
> Should be fixed with these two commits. Can you confirm?

Yes, works! Thanks a lot!

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #11 from Paul Floyd  ---
Should be fixed with these two commits. Can you confirm?

commit e5b7c187b72a55fdf329973f213688bf55e9607b (HEAD -> master, origin/master,
origin/HEAD)
Author: Paul Floyd 
Date:   Fri Dec 22 20:16:04 2023 +0100

Skip updating rodata avma and tracing when there are non-contiguous rodata
sections.

commit 171eaee3bd3f83d2d8de363711ffeb27a18ef940
Author: Paul Floyd 
Date:   Fri Dec 22 19:59:56 2023 +0100

Bug 478837 - valgrind fails to read debug info for rust binaries

mold produces binaries with non-contiguous .rodata and .rodata.xxx
sections. The code that merges their addresses can only handle 1
rodata which means only contiguous sections can be merged.

Since this now seems to be a "normal" case I've changed the warning
message to be a SYMTAB_TRACE message.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #10 from Paul Floyd  ---
OK.

--14745-- WARNING: Serious error when reading debug info
--14745-- When reading debug info from /home/paulf/foo/target/debug/foo:
--14745-- Can't make sense of .rodata section mapping
Hello, world!

  [15] .rodata
   PROGBITS f7f0  f7f0  0
   5800   0 16
   [0002]: ALLOC
  [16] .debug_gdb_scripts
   PROGBITS 00014ff0  00014ff0  0
   0022 0001  0 1
   [0032]: ALLOC, MERGE, STRINGS
  [17] .rodata.cst4
   PROGBITS 00015014  00015014  0
   004c 0004  0 4
   [0012]: ALLOC, MERGE
  [18] .rodata.cst8
   PROGBITS 00015060  00015060  0
   0040 0008  0 8
   [0012]: ALLOC, MERGE
  [19] .rodata.cst16
   PROGBITS 000150a0  000150a0  0
   0250 0010  0 16
   [0012]: ALLOC, MERGE
  [20] .rodata.cst32
   PROGBITS 000152f0  000152f0  0
   0060 0020  0 8
   [0012]: ALLOC, MERGE

I think that in all the other cases that I've seen the .rodata.xxx have always
been contiguous following .rodata.

Here there is   [16] .debug_gdb_scripts between .rodata and .rodata.xxx

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Dennis Kobert
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #9 from Dennis Kobert  ---
Yes, sorry, it only fails when using mold as a linker. I still had set `alias
cargo=mold -run cargo` in my shell config sorry about that

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Dennis Kobert
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #8 from Dennis Kobert  ---
Oh, I might be using mold as a linker, let me double check that. I'm also not
sure why my cargo version is so outdated but that shouldn't make a difference
as it works on the nix system

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #7 from Paul Floyd  ---
(In reply to Dennis Kobert from comment #6)
> I've tried it both on arch linux and nix os 23.11 with identical results. 
> rustc 1.74.1 (a28077b28 2023-12-04)
> llvm 16.0.6
> cargo 1.49.0 (d00d64df9 2020-12-05)

Arch Linux
Linux archlinux 6.6.6-arch1-1 #1 SMP PREEMPT_DYNAMIC Mon, 11 Dec 2023 11:48:23
+ x86_64 GNU/Linux
rustc 1.74.1 (a28077b28 2023-12-04) (Arch Linux rust 1:1.74.1-1)
LLD 16.0.6 (compatible with GNU linkers)
cargo 1.74.1

Still no problem.

readelf -t  ./target/debug/foo

shows just the one rodata section

  [16] .rodata
   PROGBITS 00048000  00048000  0
   6920   0 16
   [0002]: ALLOC

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Dennis Kobert
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #6 from Dennis Kobert  ---
I've tried it both on arch linux and nix os 23.11 with identical results. 
rustc 1.74.1 (a28077b28 2023-12-04)
llvm 16.0.6
cargo 1.49.0 (d00d64df9 2020-12-05)

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Dennis Kobert
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #5 from Dennis Kobert  ---
Created attachment 164366
  --> https://bugs.kde.org/attachment.cgi?id=164366&action=edit
Example binary file containing a hello world program

file foo
foo: ELF 64-bit LSB pie executable, x86-64, version 1 (SYSV), dynamically
linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 4.4.0,
BuildID[sha1]=65c7742a98850b66723b0c088d9dc219bb0a6434, with debug_info, not
stripped

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #4 from Paul Floyd  ---
Works OK for me on Fedora 39 amd64
rustc --version
rustc 1.74.1 (a28077b28 2023-12-04) (Fedora 1.74.1-1.fc39)

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #3 from Paul Floyd  ---
OK I'll try to reproduce it.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-22 Thread Dennis Kobert
https://bugs.kde.org/show_bug.cgi?id=478837

--- Comment #2 from Dennis Kobert  ---
(In reply to Paul Floyd from comment #1)
> Do you get the same problem with Valgrind built from git HEAD?

Yes, exact same behavior

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 478837] valgrind fails to read debug info for rust binaries

2023-12-21 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=478837

Paul Floyd  changed:

   What|Removed |Added

 CC||pjfl...@wanadoo.fr

--- Comment #1 from Paul Floyd  ---
Do you get the same problem with Valgrind built from git HEAD?

-- 
You are receiving this mail because:
You are watching all bug changes.