public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Lewis Hyatt <lhyatt@gmail.com>
To: David Malcolm <dmalcolm@redhat.com>
Cc: gcc-patches@gcc.gnu.org
Subject: Re: [PATCH v3 0/4] diagnostics: libcpp: Overhaul locations for _Pragma tokens
Date: Sat, 29 Jul 2023 10:27:50 -0400	[thread overview]
Message-ID: <CAA_5UQ7fgTUDmubUS7yvU=2ich73XteswYdLjkR=rr=j8gBj8Q@mail.gmail.com> (raw)
In-Reply-To: <35d4d0afb71364877f8c9755032a247f4da045cf.camel@redhat.com>

On Fri, Jul 28, 2023 at 6:22 PM David Malcolm <dmalcolm@redhat.com> wrote:
>
> On Fri, 2023-07-21 at 19:08 -0400, Lewis Hyatt wrote:
> > Hello-
> >
> > This is an update to the v2 patch series last sent in January:
> > https://gcc.gnu.org/pipermail/gcc-patches/2023-January/609473.html
> >
> > While I did not receive any feedback on the v2 patches yet, they did
> > need some
> > rebasing on top of other recent commits to input.cc, so I thought it
> > would be
> > helpful to send them again now. The patches have not otherwise
> > changed from
> > v2, and the above-linked message explains how all the patches fit in
> > with the
> > original v1 series sent last November.
> >
> > Dave, I would appreciate it very much if you could please let me know
> > what you
> > think of this approach? I feel like the diagnostics we currently
> > output for _Pragmas are worth improving. As a reminder, say for this
> > example:
> >
> > =====
> >  #define S "GCC diagnostic ignored \"oops"
> >  _Pragma(S)
> > =====
> >
> > We currently output:
> >
> > =====
> > file.cpp:2:24: warning: missing terminating " character
> >     2 | _Pragma(S)
> >       |                        ^
> > =====
> >
> > While after these patches, we would output:
> >
> > ======
> > <generated>:1:24: warning: missing terminating " character
> >     1 | GCC diagnostic ignored "oops
> >       |                        ^
> > file.cpp:2:1: note: in <_Pragma directive>
> >     2 | _Pragma(S)
> >       | ^~~~~~~
> > ======
> >
> > Thanks!
>
> Hi Lewis; sorry for not responding to the v2 patches.
>
> I've started looking at the v3 patches in detail, but I have some high-
> level questions about memory usage:
>
> Am I right in thinking that the effect of this patch is that for every
> _Pragma in the source we will create a new line_map_ordinary, and a new
> buffer for the stringified content of that _Pragma, and that these
> allocations will persist for the rest of the compilation?  (plus a
> little extra allocation within the "location_t" space from 0 to
> 0x7fffffff).
>
> It sounds like this will probably be a rounding error that won't be
> noticable in profiling, but did you attempt any such measurement of the
> memory usage before/after this patch on some real-world projects?
>
> Thanks
> Dave
>

Thanks for looking at the patches, I appreciate it whenever you have
time to get to them.

This is a fair point about the memory usage, basically it means that
each instance of a _Pragma has comparable memory footprint to a macro
definition. (In addition to the overheads you mentioned, it also
creates a macro map to generate a virtual location for the tokens, so
that it's able to output the "in expansion of _Pragma" note. That part
can be disabled with -ftrack-macro-expansion=0 at least.)

I had the sense that _Pragma isn't used often enough for that to be a
problem, but agreed it is worth checking. (I really hope this memory
usage isn't an issue since there are also numerous PRs complaining
about 32-bit limitations in location tracking, that make it tempting
to explore 64-bit line maps or some other option someday too.)

I tried one thing now, wxWidgets uses a lot of diagnostic pragmas
wrapped up inside macros that use _Pragma. (See
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55578). The testsuite
contains a file allheaders.cpp which includes the whole library, so I
tried compiling this into a pch, which I believe measures the entire
memory footprint including the ordinary and macro line maps and the
_Pragma strings. The resulting PCH sizes were:

279000173 bytes before the changes
279491345 bytes after the changes

So 0.1% bigger. Happy to check other projects too, do you have any
standard gotos? Maybe firefox or something I take it.

I see your other response on patch #1, I am thinking about that and
will reply later. Thanks again!

-Lewis

  reply	other threads:[~2023-07-29 14:28 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-07-21 23:08 Lewis Hyatt
2023-07-21 23:08 ` [PATCH v3 1/4] diagnostics: libcpp: Add LC_GEN linemaps to support in-memory buffers Lewis Hyatt
2023-07-28 22:58   ` David Malcolm
2023-07-31 22:39     ` Lewis Hyatt
2023-08-09 22:14       ` [PATCH v4 0/8] diagnostics: libcpp: Overhaul locations for _Pragma tokens Lewis Hyatt
2023-08-09 22:14         ` [PATCH v4 1/8] libcpp: Add LC_GEN linemaps to support in-memory buffers Lewis Hyatt
2023-08-11 22:45           ` David Malcolm
2023-08-13 20:18             ` Lewis Hyatt
2023-08-09 22:14         ` [PATCH v4 2/8] libcpp: diagnostics: Support generated data in expanded locations Lewis Hyatt
2023-08-11 23:02           ` David Malcolm
2023-08-14 21:41             ` Lewis Hyatt
2023-08-09 22:14         ` [PATCH v4 3/8] diagnostics: Refactor class file_cache_slot Lewis Hyatt
2023-08-15 15:43           ` David Malcolm
2023-08-15 17:58             ` Lewis Hyatt
2023-08-15 19:39               ` David Malcolm
2023-08-23 21:22                 ` Lewis Hyatt
2023-08-09 22:14         ` [PATCH v4 4/8] diagnostics: Support obtaining source code lines from generated data buffers Lewis Hyatt
2023-08-15 16:15           ` David Malcolm
2023-08-15 18:15             ` Lewis Hyatt
2023-08-15 19:46               ` David Malcolm
2023-08-15 20:08                 ` Lewis Hyatt
2023-08-23 19:41                   ` Lewis Hyatt
2023-08-09 22:14         ` [PATCH v4 5/8] diagnostics: Support testing generated data in input.cc selftests Lewis Hyatt
2023-08-15 16:27           ` David Malcolm
2023-08-09 22:14         ` [PATCH v4 6/8] diagnostics: Full support for generated data locations Lewis Hyatt
2023-08-15 16:39           ` David Malcolm
2023-08-09 22:14         ` [PATCH v4 7/8] diagnostics: libcpp: Assign real locations to the tokens inside _Pragma strings Lewis Hyatt
2023-08-09 22:14         ` [PATCH v4 8/8] diagnostics: Support generated data locations in SARIF output Lewis Hyatt
2023-08-15 17:04           ` David Malcolm
2023-08-15 17:51             ` Lewis Hyatt
2023-07-21 23:08 ` [PATCH v3 2/4] diagnostics: Handle generated data locations in edit_context Lewis Hyatt
2023-07-21 23:08 ` [PATCH v3 3/4] diagnostics: libcpp: Assign real locations to the tokens inside _Pragma strings Lewis Hyatt
2023-07-21 23:08 ` [PATCH v3 4/4] diagnostics: Support generated data locations in SARIF output Lewis Hyatt
2023-07-28 22:22 ` [PATCH v3 0/4] diagnostics: libcpp: Overhaul locations for _Pragma tokens David Malcolm
2023-07-29 14:27   ` Lewis Hyatt [this message]
2023-07-29 16:03     ` David Malcolm

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='CAA_5UQ7fgTUDmubUS7yvU=2ich73XteswYdLjkR=rr=j8gBj8Q@mail.gmail.com' \
    --to=lhyatt@gmail.com \
    --cc=dmalcolm@redhat.com \
    --cc=gcc-patches@gcc.gnu.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).