From: Bruno Larsen via Gdb-patches <gdb-patches@sourceware.org>
To: Simon Marchi <simark@simark.ca>, gdb-patches@sourceware.org
Subject: Re: [PATCH] gdb: make "start" breakpoint inferior-specific
Date: Mon, 7 Nov 2022 09:14:26 +0100 [thread overview]
Message-ID: <2cd08b65-12e0-3763-5f41-cc1caad4f554@redhat.com> (raw)
In-Reply-To: <a7ad6c5d-6fef-07de-20d3-112a4fcbf833@simark.ca>
On 04/11/2022 17:52, Simon Marchi wrote:
> On 8/31/22 10:03, Bruno Larsen via Gdb-patches wrote:
>> On 04/08/2022 19:40, Simon Marchi via Gdb-patches wrote:
>>> I saw this failure on a CI:
>>>
>>> (gdb) add-inferior
>>> [New inferior 2]
>>> Added inferior 2
>>> (gdb) PASS: gdb.threads/vfork-multi-inferior.exp: method=non-stop: add-inferior
>>> inferior 2
>>> [Switching to inferior 2 [<null>] (<noexec>)]
>>> (gdb) PASS: gdb.threads/vfork-multi-inferior.exp: method=non-stop: inferior 2
>>> kill
>>> The program is not being run.
>>> (gdb) file /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/tmp/tmp.GYATAXR8Ku/gdb/testsuite/outputs/gdb.threads/vfork-multi-inferior/vfork-multi-inferior-sleep
>>> Reading symbols from /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/tmp/tmp.GYATAXR8Ku/gdb/testsuite/outputs/gdb.threads/vfork-multi-inferior/vfork-multi-inferior-sleep...
>>> (gdb) run &
>>> Starting program: /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/tmp/tmp.GYATAXR8Ku/gdb/testsuite/outputs/gdb.threads/vfork-multi-inferior/vfork-multi-inferior-sleep
>>> (gdb) PASS: gdb.threads/vfork-multi-inferior.exp: method=non-stop: run inferior 2
>>> inferior 1
>>> [Switching to inferior 1 [<null>] (<noexec>)]
>>> (gdb) PASS: gdb.threads/vfork-multi-inferior.exp: method=non-stop: inferior 1
>>> kill
>>> The program is not being run.
>>> (gdb) file /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/tmp/tmp.GYATAXR8Ku/gdb/testsuite/outputs/gdb.threads/vfork-multi-inferior/vfork-multi-inferior
>>> Reading symbols from /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/tmp/tmp.GYATAXR8Ku/gdb/testsuite/outputs/gdb.threads/vfork-multi-inferior/vfork-multi-inferior...
>>> (gdb) break should_break_here
>>> Breakpoint 1 at 0x11b1: file /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/src/binutils-gdb/gdb/testsuite/gdb.threads/vfork-multi-inferior.c, line 25.
>>> (gdb) PASS: gdb.threads/vfork-multi-inferior.exp: method=non-stop: break should_break_here
>>> [Thread debugging using libthread_db enabled]
>>> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
>>> start
>>> Temporary breakpoint 2 at 0x11c0: -qualified main. (2 locations)
>>> Starting program: /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/tmp/tmp.GYATAXR8Ku/gdb/testsuite/outputs/gdb.threads/vfork-multi-inferior/vfork-multi-inferior
>>> [Thread debugging using libthread_db enabled]
>>> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
>>>
>>> Thread 2.1 "vfork-multi-inf" hit Temporary breakpoint 2, main () at /home/jenkins/workspace/binutils-gdb_master_linuxbuild/platform/jammy-amd64/target_board/unix/src/binutils-gdb/gdb/testsuite/gdb.threads/vfork-multi-inferior-sleep.c:23
>>> 23 sleep (30);
>>> (gdb) FAIL: gdb.threads/vfork-multi-inferior.exp: method=non-stop: start inferior 1
>>>
>>> What happens is:
>>>
>>> 1. We start inferior 2 with "run&", it runs very slowly, takes time to
>>> get to main
>>> 2. We switch to inferior 1, and run "start"
>>> 3. The temporary breakpoint inserted by "start" applies to all inferiors
>>> 4. Inferior 2 hits that breakpoint and GDB reports that hit
>>>
>>> To avoid this, breakpoints inserted by "start" should be
>>> inferior-specific. However, we don't have a nice way to make
>>> inferior-specific breakpoints yet. It's possible to make
>>> pspace-specific breakpoints (for example how the internal_breakpoint
>>> constructor does) by creating a symtab_and_line manually. However,
>>> inferiors can share program spaces (usually on particular embedded
>>> targets), so we could have a situation where two inferiors run the same
>>> code in the same program space. In that case, it would just not be
>>> possible to insert a breakpoint in one inferior but not the other.
>>>
>>> A simple solution that should work all the time is to add a condition to
>>> the breakpoint inserted by "start", to check the inferior reporting the
>>> hit is the expected one. This is what this patch implements.
>>>
>>> Add a test that does:
>>>
>>> - start a background inferior 1 that calls main in a loop
>>> - start aninferior 2 with the "start" command
>>> - validate that we hit the breakpoint in inferior 2
>>>
>>> Without the fix, we hit the breakpoint in inferior 1 pretty much all the
>>> time.
>>>
>>> Change-Id: Ib0148498a476bfa634ed62353c95f163623c686a
>> Hi Simon,
>>
>> This change makes a lot of sense, I just wonder if it would make sense to make breakpoints inferior-specific. I could see a situation where someone would run the same program twice, one with an ok input while the other has a buggy input for instance, to see the difference between them, and inferior-specific breakpoints could come in handy in a situation like this.
>>
>> If you think it would be too much work for a usecase that is too niche, this patch is pretty straight forward, I'd suggest you approve your patch.
> I certainly don't want to add a new user-facing feature just for this
> fix, as that's a lot of work (figuring out the syntax, doc, NEWS, tests,
> etc). We could enhance the internal API to allow passing an inferior to
> make a breakpoint inferior-specific, but I don't think it should be a
> prerequisite for this fix. It would end up doing functionally
> identical.
What I had in mind is closer to your second idea, just internal API
changes to keep track of the inferior, though maybe I just hadn't
thought through the suggestion, since letting users add breakpoint to
other inferiors makes sense given the use case I mentioned above.
This is a long winded way of saying I agree that this is unreasonable
feature creep for a functionally identical end result for now, so the
patch LGTM.
Reviewed-By: Bruno Larsen <blarsen@redhat.com>
Cheers,
Bruno
>
> Would you like to to add your Reviewed-By on this patch?
>
> Simon
>
next prev parent reply other threads:[~2022-11-07 8:15 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-04 17:40 Simon Marchi via Gdb-patches
2022-08-17 17:56 ` Simon Marchi via Gdb-patches
2022-08-31 14:03 ` Bruno Larsen via Gdb-patches
2022-11-04 16:52 ` Simon Marchi via Gdb-patches
2022-11-07 8:14 ` Bruno Larsen via Gdb-patches [this message]
2022-11-08 17:24 ` Tom Tromey
2022-09-01 10:42 ` Andrew Burgess via Gdb-patches
2022-11-04 17:24 ` Simon Marchi via Gdb-patches
[not found] ` <8735asb7cj.fsf@redhat.com>
2022-11-09 13:19 ` Simon Marchi via Gdb-patches
2022-11-08 19:43 ` Pedro Alves
2022-11-08 20:14 ` Simon Marchi via Gdb-patches
2022-11-08 21:09 ` Pedro Alves
2022-11-08 21:20 ` [PATCH v2] " Simon Marchi via Gdb-patches
2022-11-10 16:45 ` Pedro Alves
2022-11-10 17:33 ` Simon Marchi via Gdb-patches
2022-11-10 17:36 ` Simon Marchi via Gdb-patches
2022-11-10 17:47 ` Pedro Alves
2022-11-10 17:53 ` Simon Marchi via Gdb-patches
2022-11-11 12:37 ` Tom de Vries via Gdb-patches
2022-11-11 13:53 ` Simon Marchi via Gdb-patches
2022-11-11 15:21 ` Tom de Vries via Gdb-patches
2022-11-11 19:03 ` Simon Marchi via Gdb-patches
2022-11-12 10:43 ` Tom de Vries via Gdb-patches
2022-11-14 11:29 ` Tom de Vries via Gdb-patches
2022-11-14 13:19 ` Simon Marchi via Gdb-patches
2022-11-14 14:18 ` Tom de Vries via Gdb-patches
2022-11-16 16:22 ` Tom Tromey
2022-11-16 16:26 ` Simon Marchi via Gdb-patches
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=2cd08b65-12e0-3763-5f41-cc1caad4f554@redhat.com \
--to=gdb-patches@sourceware.org \
--cc=blarsen@redhat.com \
--cc=simark@simark.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