* [Bug symtab/16405] backtrace locks up on Fedora 20
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
@ 2014-01-07 20:30 ` jan.kratochvil at redhat dot com
2014-01-10 20:49 ` jan.kratochvil at redhat dot com
` (7 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: jan.kratochvil at redhat dot com @ 2014-01-07 20:30 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
Jan Kratochvil <jan.kratochvil at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jan.kratochvil at redhat dot com
Summary|call-sites lock up |backtrace locks up on
|backtrace on Fedora 20 |Fedora 20
--- Comment #1 from Jan Kratochvil <jan.kratochvil at redhat dot com> ---
call-sites are really innocent, it is DWARF expansion problem in general.
I have patched out call-sites in GDB code and the problem remains.
#0 in start_subfile
#1 in dwarf2_start_subfile
#2 in dwarf_decode_lines
#3 in handle_DW_AT_stmt_list
#4 in read_file_scope
#5 in process_die
#6 in process_full_comp_unit
#7 in process_queue
#8 in dw2_do_instantiate_symtab
#9 in dw2_instantiate_symtab
#10 in dw2_find_pc_sect_symtab
#11 in find_pc_sect_symtab
#12 in blockvector_for_pc_sect
#13 in block_for_pc_sect
#14 in block_for_pc
#15 in inline_frame_sniffer
#16 in frame_unwind_find_by_frame
#17 in compute_frame_id
#18 in get_prev_frame_if_no_cycle
#19 in get_prev_frame_1
#20 in get_prev_frame
#21 in value_of_dwarf_reg_entry
#22 in value_of_dwarf_block_entry
#23 in loclist_read_variable_at_entry
#24 in read_frame_arg
#25 in print_frame_args
#26 in print_frame
#27 in print_frame_info
#28 in backtrace_command_1
#29 in backtrace_command
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace locks up on Fedora 20
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
2014-01-07 20:30 ` [Bug symtab/16405] backtrace locks up " jan.kratochvil at redhat dot com
@ 2014-01-10 20:49 ` jan.kratochvil at redhat dot com
2014-01-10 20:50 ` [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m jan.kratochvil at redhat dot com
` (6 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: jan.kratochvil at redhat dot com @ 2014-01-10 20:49 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
Jan Kratochvil <jan.kratochvil at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |ASSIGNED
Assignee|unassigned at sourceware dot org |jan.kratochvil at redhat dot com
--- Comment #2 from Jan Kratochvil <jan.kratochvil at redhat dot com> ---
Created attachment 7347
--> https://sourceware.org/bugzilla/attachment.cgi?id=7347&action=edit
draft patch
This patch:
bt
Command execution time: 0.472000 (cpu), 0.539249 (wall)
Space used: 380514304 (+44015616 for this command)
#symtabs: 30964 (+2457), #primary symtabs: 18 (+7), #blocks: 7877 (+7852)
+
thread apply all bt
Command execution time: 0.496000 (cpu), 0.500745 (wall)
Space used: 408891392 (+28377088 for this command)
#symtabs: 32294 (+1330), #primary symtabs: 31 (+13), #blocks: 16467 (+8590)
FSF GDB HEAD:
bt
Command execution time: 153.667000 (cpu), 154.234038 (wall)
Space used: 1283444736 (+946941952 for this command)
#symtabs: 4373195 (+4344688), #primary symtabs: 2376 (+2365), #blocks: 12593
(+12568)
+
thread apply all bt
Command execution time: 104.489000 (cpu), 104.961148 (wall)
Space used: 1768660992 (+485216256 for this command)
#symtabs: 6662638 (+2289443), #primary symtabs: 3673 (+1297), #blocks: 23751
(+11158)
But
thread apply all bt full
is still slow. One may omit DW_TAG_imported_unit even in this case thanks to
.gdb_index (if it points to the right CU, maybe it is not). This still affects
for example ABRT.
gdb -p `pidof soffice.bin` -ex 'set pagination off' -ex 'maintenance set
per-command space' -ex 'maintenance set per-command symtab' -ex 'maintenance
set per-command time'
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
2014-01-07 20:30 ` [Bug symtab/16405] backtrace locks up " jan.kratochvil at redhat dot com
2014-01-10 20:49 ` jan.kratochvil at redhat dot com
@ 2014-01-10 20:50 ` jan.kratochvil at redhat dot com
2014-05-05 21:42 ` mjw at redhat dot com
` (5 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: jan.kratochvil at redhat dot com @ 2014-01-10 20:50 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
Jan Kratochvil <jan.kratochvil at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Summary|backtrace locks up on |backtrace takes GBs and
|Fedora 20 |minutes with dwz -m
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
` (2 preceding siblings ...)
2014-01-10 20:50 ` [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m jan.kratochvil at redhat dot com
@ 2014-05-05 21:42 ` mjw at redhat dot com
2014-10-01 18:24 ` jan.kratochvil at redhat dot com
` (4 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: mjw at redhat dot com @ 2014-05-05 21:42 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
Mark Wielaard <mjw at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |mjw at redhat dot com
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
` (3 preceding siblings ...)
2014-05-05 21:42 ` mjw at redhat dot com
@ 2014-10-01 18:24 ` jan.kratochvil at redhat dot com
2014-10-01 19:49 ` jan.kratochvil at redhat dot com
` (3 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: jan.kratochvil at redhat dot com @ 2014-10-01 18:24 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
Jan Kratochvil <jan.kratochvil at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #7347|0 |1
is obsolete| |
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
` (4 preceding siblings ...)
2014-10-01 18:24 ` jan.kratochvil at redhat dot com
@ 2014-10-01 19:49 ` jan.kratochvil at redhat dot com
2014-10-06 17:58 ` sergiodj at redhat dot com
` (2 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: jan.kratochvil at redhat dot com @ 2014-10-01 19:49 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
--- Comment #3 from Jan Kratochvil <jan.kratochvil at redhat dot com> ---
[patch] Fix 100x slowdown regression on DWZ files
https://sourceware.org/ml/gdb-patches/2014-10/msg00020.html
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
` (5 preceding siblings ...)
2014-10-01 19:49 ` jan.kratochvil at redhat dot com
@ 2014-10-06 17:58 ` sergiodj at redhat dot com
2015-01-24 15:03 ` jan.kratochvil at redhat dot com
2015-01-25 15:33 ` jan.kratochvil at redhat dot com
8 siblings, 0 replies; 10+ messages in thread
From: sergiodj at redhat dot com @ 2014-10-06 17:58 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
Sergio Durigan Junior <sergiodj at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |sergiodj at redhat dot com
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
` (6 preceding siblings ...)
2014-10-06 17:58 ` sergiodj at redhat dot com
@ 2015-01-24 15:03 ` jan.kratochvil at redhat dot com
2015-01-25 15:33 ` jan.kratochvil at redhat dot com
8 siblings, 0 replies; 10+ messages in thread
From: jan.kratochvil at redhat dot com @ 2015-01-24 15:03 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
Jan Kratochvil <jan.kratochvil at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|ASSIGNED |RESOLVED
Resolution|--- |FIXED
--- Comment #4 from Jan Kratochvil <jan.kratochvil at redhat dot com> ---
527f3840e1af8bc2e3173922ddae15d0021ed9b1 master
0325b9bae8037aaa69cdfc8aa43a46678474f261 7.9
time gdb -batch -q -p `pidof soffice.bin` -ex 'thread apply all bt full'|wc -l
174
real 0m6.805s
time gdb -batch -q -p `pidof soffice.bin` -ex 'thread apply all bt'|wc -l
61
real 0m3.635s
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Bug symtab/16405] backtrace takes GBs and minutes with dwz -m
2014-01-07 18:16 [Bug symtab/16405] New: call-sites lock up backtrace on Fedora 20 jan.kratochvil at redhat dot com
` (7 preceding siblings ...)
2015-01-24 15:03 ` jan.kratochvil at redhat dot com
@ 2015-01-25 15:33 ` jan.kratochvil at redhat dot com
8 siblings, 0 replies; 10+ messages in thread
From: jan.kratochvil at redhat dot com @ 2015-01-25 15:33 UTC (permalink / raw)
To: gdb-prs
https://sourceware.org/bugzilla/show_bug.cgi?id=16405
--- Comment #5 from Jan Kratochvil <jan.kratochvil at redhat dot com> ---
The Comment 4 benchmark is BTW bogus as it did not have expanded CUs.
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 10+ messages in thread