#583 GDB 20040521 callstack 10 times slowdown regression

[Eugene]

In GDB 20040521 commands
"-stack-list-frames" "-stack-list-arguments 0"
are 10-20 times slower and in some cases are executed
in seconds. (Up to 30 seconds with stack depth 16 on
P4 2.4).

Is it a build regression or developers regression?
What can be done here?

Was mingw build gdb 20040521 compiled with
optimizations?
On which version of gcc?

[Chris Johns]

Logged In: YES
user_id=197074

Does this still cause a problem with gdb-6.3-2 ?

[Chris Johns]

Logged In: YES
user_id=197074

Does this still cause a problem with gdb-6.3-2 ?

[Eugene]

Logged In: YES
user_id=1022506

I'll test it when I'll have the possibility to do it.

[Eugene]

Logged In: YES
user_id=1022506
Originator: YES

As for me the case still exists on the latest gdb mingw edition.
FYI Gdb is still too buggy when it comes to the need to get a callstack in the code which consists of two dll's runtime - the msvc compiled one calls some function from the gdb compiled dll.
All the problems are in such situation (Slowdown and constant gpf's getting callstack) .

[Earnie Boyd]

If this issue still exists please report it to gdb bugs queue.