From: "vries at gcc dot gnu.org" <sourceware-bugzilla@sourceware.org>
To: dwz@sourceware.org
Subject: [Bug default/25291] New: create_import_tree phase 2 slowness
Date: Tue, 01 Jan 2019 00:00:00 -0000 [thread overview]
Message-ID: <bug-25291-11298@http.sourceware.org/bugzilla/> (raw)
https://sourceware.org/bugzilla/show_bug.cgi?id=25291
Bug ID: 25291
Summary: create_import_tree phase 2 slowness
Product: dwz
Version: unspecified
Status: NEW
Severity: enhancement
Priority: P2
Component: default
Assignee: nobody at sourceware dot org
Reporter: vries at gcc dot gnu.org
CC: dwz at sourceware dot org
Target Milestone: ---
With the odr patches series, but with odr disabled we have for the clang-10
benchmark:
...
$ dwz clang-10.debug -o 1 -lnone -Lnone --devel-progress --no-odr
read_debug_info .debug_info
user: 59.70
sys : 3.29
partition_dups
user: 1.20
sys : 0.01
partition_dups qsort
user: 1.64
sys : 0.00
partition_dups after qsort
user: 1.64
sys : 0.35
create_import_tree phase 1
user: 0.44
sys : 0.01
create_import_tree phase 2
user: 1.35
sys : 0.00
create_import_tree phase 3
user: 1.26
sys : 0.00
create_import_tree phase 4 (create partial units)
user: 0.24
sys : 0.02
read_debug_info .debug_types
user: 1.36
sys : 0.00
compute_abbrevs
user: 37.41
sys : 0.31
write_abbrev
user: 0.09
sys : 0.00
write_info
user: 12.48
sys : 0.86
write_loc
user: 3.63
sys : 8.69
write_types
user: 0.00
sys : 0.00
write_gdb_index
user: 0.17
sys : 0.28
write_dso
user: 0.18
sys : 2.32
...
And with --odr --odr-mode=basic:
...
$ dwz clang-10.debug -o 1 -lnone -Lnone --devel-progress --odr --odr-mode=basic
read_debug_info .debug_info
user: 46.39
sys : 3.36
partition_dups
user: 2.14
sys : 0.01
partition_dups split_dups
user: 99.34
sys : 0.51
partition_dups qsort
user: 4.82
sys : 0.00
partition_dups after qsort
user: 3.76
sys : 0.23
create_import_tree phase 1
user: 1.02
sys : 0.02
create_import_tree phase 2
user: 48.87
sys : 0.00
create_import_tree phase 3
user: 11.73
sys : 0.00
create_import_tree phase 4 (create partial units)
user: 2.78
sys : 0.05
read_debug_info .debug_types
user: 1.40
sys : 0.00
compute_abbrevs
user: 46.84
sys : 0.35
write_abbrev
user: 0.08
sys : 0.00
write_info
user: 8.48
sys : 0.97
write_loc
user: 0.82
sys : 0.58
write_types
user: 0.00
sys : 0.00
write_gdb_index
user: 0.21
sys : 0.48
write_dso
user: 0.07
sys : 1.34
...
The largest addition in time is the odr phase split_dups, which is a known
issue.
But the second largest addition is in create_import_tree phase 2, from 1.35s to
48.87. It would be good to understand how this increase happens, and if we can
reduce it.
--
You are receiving this mail because:
You are on the CC list for the bug.
next reply other threads:[~2019-12-18 8:30 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-01-01 0:00 vries at gcc dot gnu.org [this message]
2019-01-01 0:00 ` [Bug default/25291] " vries at gcc dot gnu.org
2021-03-03 9:12 ` vries at gcc dot gnu.org
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=bug-25291-11298@http.sourceware.org/bugzilla/ \
--to=sourceware-bugzilla@sourceware.org \
--cc=dwz@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).