From: Tom de Vries <tdevries@suse.de>
To: Simon Marchi <simon.marchi@polymtl.ca>, gdb-patches@sourceware.org
Subject: Re: [PATCH][gdb/testsuite] Ignore DEBUGINFOD_URLS
Date: Wed, 31 Mar 2021 15:21:13 +0200 [thread overview]
Message-ID: <852e5449-00d5-001d-77d3-7e25e702428c@suse.de> (raw)
In-Reply-To: <ac1a1d14-9bc0-9837-acfe-30df7ec09e42@polymtl.ca>
On 3/31/21 3:04 PM, Simon Marchi wrote:
> On 2021-03-31 8:33 a.m., Tom de Vries wrote:> Hi,
>>
>> On openSUSE Tumbleweed, DEBUGINFOD_URLS is now defined by default:
>> ...
>> $ echo $DEBUGINFOD_URLS
>> https://debuginfod.opensuse.org/
>> ...
>>
>> With DEBUGINFOD_URLS defined we run into:
>> ...
>> FAIL: gdb.mi/mi-sym-info.exp: List all functions from debug information only \
>> (timeout)
>> ...
>> as reported in PR27667.
>>
>> There's a latency of ~0.5s per request, which is ok-ish for interactive usage.
>> But the symbol-info-functions command ends up issuing 21 source requests,
>> which means we easily run into the 10s timeout.
>>
>> Fix this by unsetting DEBUGINFOD_URLS in default_gdb_init.
>>
>> Any comments?
>>
>> Thanks,
>> - Tom
>>
>> [gdb/testsuite] Ignore DEBUGINFOD_URLS
>>
>> gdb/testsuite/ChangeLog:
>>
>> 2021-03-31 Tom de Vries <tdevries@suse.de>
>>
>> PR testsuite/27667
>> * lib/gdb.exp (default_gdb_init): Unset DEBUGINFOD_URLS.
>>
>> ---
>> gdb/testsuite/lib/gdb.exp | 4 ++++
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
>> index 73fea3a104d..d18aec648d9 100644
>> --- a/gdb/testsuite/lib/gdb.exp
>> +++ b/gdb/testsuite/lib/gdb.exp
>> @@ -5270,6 +5270,10 @@ proc default_gdb_init { test_file_name } {
>> # tests.
>> setenv TERM "dumb"
>>
>> + # If DEBUGINFOD_URLS is set, gdb will try to download sources and
>> + # debug info for f.i. system libraries. Prevent this.
>> + unsetenv DEBUGINFOD_URLS
>> +
>> # Ensure that GDBHISTFILE and GDBHISTSIZE are removed from the
>> # environment, we don't want these modifications to the history
>> # settings.
>>
>
> I think this is OK, we want to reduce external influence on the tests as
> much as possible.
>
> However, I could imagine we could have a "debuginfod" board file, which
> runs all tests with a controlled debuginfod server it spawns itself, and
> it would set DEBUGINFOD_URLS accordingly. It could override gdb_compile
> to feed all compiled binaries to the debuginfod server, something like
> that.
I also thought about that possibility.
Anyway, such a board file could define something like:
...
proc gdb_init { args } {
set res [default_gdb_init {*}$args]
setenv DEBUGINFOD_URLS "..."
return $res
}
...
Committed, with the test for info exists env(DEBUGINFOD_URLS) added.
Thanks,
- Tom
prev parent reply other threads:[~2021-03-31 13:21 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-31 12:33 Tom de Vries
2021-03-31 13:04 ` Simon Marchi via Gdb-patches
2021-03-31 13:21 ` Tom de Vries [this message]
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=852e5449-00d5-001d-77d3-7e25e702428c@suse.de \
--to=tdevries@suse.de \
--cc=gdb-patches@sourceware.org \
--cc=simon.marchi@polymtl.ca \
/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