public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Andrew Burgess <aburgess@redhat.com>
To: Tom de Vries <tdevries@suse.de>, gdb-patches@sourceware.org
Subject: Re: [PATCH 1/3] [gdb/testsuite] Mention DEBUGINFOD_VERBOSE in gdb.debuginfod test-cases
Date: Mon, 27 May 2024 22:40:51 +0100	[thread overview]
Message-ID: <87o78r9bmk.fsf@redhat.com> (raw)
In-Reply-To: <20240527141808.31563-1-tdevries@suse.de>

Tom de Vries <tdevries@suse.de> writes:

> In proc start_debuginfod we start debuginfod with a fair amount of verbosity:
> ...
> 	spawn debuginfod -vvvv -d $db -p $port -F $debugdir
> ...
> which gives a lot of information on startup, but none or not much after that.
>
> There's another way to increase verbosity of debuginfod: through the
> environment variable DEBUGINFOD_VERBOSE, which gives us the following
> additional information (edited for readability):
> ...
> (gdb) file fetch_src_and_symbols^M
>  Reading symbols from fetch_src_and_symbols...^M
> +debuginfod_find_debuginfo $hex^M
> +server urls "http://127.0.0.1:8000"^M
> +checking build-id^M
> +checking cache dir $client_cache^M
> +using timeout 10^M
> +init server 0 http://127.0.0.1:8000^M
> +url 0 http://127.0.0.1:8000/buildid/$hex/debuginfo^M
> +query 1 urls in parallel^M
>  Downloading separate debug info for $fetch_src_and_symbols...^M
> +server response No error^M
> +got file from server^M
> +found $client_cache/$hex/debuginfo (fd=18)^M
>  Reading symbols from $client_cache/$hex/debuginfo...^M
> (gdb)
> ...
>
> IMO this is too verbose to enable by default, but it's good to at least
> mention the option explicitly in the test-cases.
>
> Tested by enabling it in the 3 gdb.debuginfod test-cases.  This generates 2
> FAILs, but I don't think it's worthwhile to address these.
>
> Tested on x86_64-linux.
> ---
>  gdb/testsuite/gdb.debuginfod/build-id-no-debug-warning.exp | 3 +++
>  gdb/testsuite/gdb.debuginfod/crc_mismatch.exp              | 3 +++
>  gdb/testsuite/gdb.debuginfod/fetch_src_and_symbols.exp     | 3 +++
>  3 files changed, 9 insertions(+)
>
> diff --git a/gdb/testsuite/gdb.debuginfod/build-id-no-debug-warning.exp b/gdb/testsuite/gdb.debuginfod/build-id-no-debug-warning.exp
> index b86622543ef..25d800a378b 100644
> --- a/gdb/testsuite/gdb.debuginfod/build-id-no-debug-warning.exp
> +++ b/gdb/testsuite/gdb.debuginfod/build-id-no-debug-warning.exp
> @@ -127,6 +127,9 @@ prepare_for_debuginfod cache db
>  proc_with_prefix local_debuginfod { } {
>      global db debuginfod_debugdir cache build_id_debug_file
>  
> +    # Uncomment to increase verbosity of debuginfod when handling query.
> +    #setenv DEBUGINFOD_VERBOSE 1
> +

I'm not a huge fan of leaving comment out code like this.  I have some
patches pending that add more debuginfod tests.  Am I expected to add
this same commented out code before every call to start_debuginfod?

Some alternative suggestions:

  - Add this comment to start_debuginfod's comment, as in: "If
    additional debugging is required try 'setenv DEBUGINFOD_VERBOSE 1'."

  - We could document this in testsuite/README as a hint for debugging
    debuginfod tests.

  - We could add an 'args' list argument to start_debuginfod, and accept
    a 'verbose=1' argument within args.

Hopefully you might find one of these an acceptable alternative...

thanks,
Andrew


>      set url [start_debuginfod $db $debuginfod_debugdir]
>      if {$url eq ""} {
>  	unresolved "failed to start debuginfod server"
> diff --git a/gdb/testsuite/gdb.debuginfod/crc_mismatch.exp b/gdb/testsuite/gdb.debuginfod/crc_mismatch.exp
> index 017ef132573..27069b580dc 100644
> --- a/gdb/testsuite/gdb.debuginfod/crc_mismatch.exp
> +++ b/gdb/testsuite/gdb.debuginfod/crc_mismatch.exp
> @@ -99,6 +99,9 @@ proc_with_prefix local_debuginfod { } {
>      global binfile db debugdir cache
>      set escapedobjdirsubdir [string_to_regexp [standard_output_file {}]]
>  
> +    # Uncomment to increase verbosity of debuginfod when handling query.
> +    #setenv DEBUGINFOD_VERBOSE 1
> +
>      set url [start_debuginfod $db $debugdir]
>      if {$url eq ""} {
>  	unresolved "failed to start debuginfod server"
> diff --git a/gdb/testsuite/gdb.debuginfod/fetch_src_and_symbols.exp b/gdb/testsuite/gdb.debuginfod/fetch_src_and_symbols.exp
> index 401af0df0d2..0fb752ff6bd 100644
> --- a/gdb/testsuite/gdb.debuginfod/fetch_src_and_symbols.exp
> +++ b/gdb/testsuite/gdb.debuginfod/fetch_src_and_symbols.exp
> @@ -209,6 +209,9 @@ proc disable_delete_breakpoints {} {
>  proc_with_prefix local_url { } {
>      global binfile outputdir debugdir db
>  
> +    # Uncomment to increase verbosity of debuginfod when handling query.
> +    #setenv DEBUGINFOD_VERBOSE 1
> +
>      set url [start_debuginfod $db $debugdir]
>      if { $url == "" } {
>  	unresolved "failed to start debuginfod server"
>
> base-commit: 4250085217f2011335257fd3291cb50c939e9746
> -- 
> 2.35.3


      parent reply	other threads:[~2024-05-27 21:40 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-27 14:18 Tom de Vries
2024-05-27 14:18 ` [PATCH 2/3] [gdb/symtab] Add debuginfod verbose level 2 Tom de Vries
2024-05-27 21:48   ` Andrew Burgess
2024-05-27 14:18 ` [PATCH 3/3] [gdb/symtab] Note success at " Tom de Vries
2024-05-27 21:49   ` Andrew Burgess
2024-05-27 21:40 ` Andrew Burgess [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=87o78r9bmk.fsf@redhat.com \
    --to=aburgess@redhat.com \
    --cc=gdb-patches@sourceware.org \
    --cc=tdevries@suse.de \
    /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).