From: Austin Clements <amdragon@MIT.EDU>
To: Mark Walters <markwalters1009@gmail.com>
Cc: notmuch@notmuchmail.org
Subject: Re: [PATCH 2/3] new: Handle fatal errors in remove_filename and _remove_directory
Date: Sun, 22 Apr 2012 00:21:05 -0400 [thread overview]
Message-ID: <20120422042105.GB17810@mit.edu> (raw)
In-Reply-To: <87wr5f4qci.fsf@qmul.ac.uk>
Quoth Mark Walters on Apr 16 at 5:02 pm:
> On Mon, 27 Feb 2012, Austin Clements <amdragon@MIT.EDU> wrote:
> > Previously such errors were simply ignored. Now they cause an
> > immediate cleanup and abort.
>
> This one looks fine except for a minor query.
>
> > ---
> > notmuch-new.c | 24 ++++++++++++++++++------
> > 1 files changed, 18 insertions(+), 6 deletions(-)
> >
> > diff --git a/notmuch-new.c b/notmuch-new.c
> > index bd9786a..0cbd479 100644
> > --- a/notmuch-new.c
> > +++ b/notmuch-new.c
> > @@ -780,8 +780,10 @@ remove_filename (notmuch_database_t *notmuch,
> > add_files_state->renamed_messages++;
> > if (add_files_state->synchronize_flags == TRUE)
> > notmuch_message_maildir_flags_to_tags (message);
> > - } else
> > + status = NOTMUCH_STATUS_SUCCESS;
> > + } else if (status == NOTMUCH_STATUS_SUCCESS) {
> > add_files_state->removed_messages++;
> > + }
> > notmuch_message_destroy (message);
> > notmuch_database_end_atomic (notmuch);
> > return status;
> > @@ -789,12 +791,13 @@ remove_filename (notmuch_database_t *notmuch,
> >
> > /* Recursively remove all filenames from the database referring to
> > * 'path' (or to any of its children). */
> > -static void
> > +static notmuch_status_t
> > _remove_directory (void *ctx,
> > notmuch_database_t *notmuch,
> > const char *path,
> > add_files_state_t *add_files_state)
> > {
> > + notmuch_status_t status;
> > notmuch_directory_t *directory;
> > notmuch_filenames_t *files, *subdirs;
> > char *absolute;
> > @@ -807,8 +810,10 @@ _remove_directory (void *ctx,
> > {
> > absolute = talloc_asprintf (ctx, "%s/%s", path,
> > notmuch_filenames_get (files));
> > - remove_filename (notmuch, absolute, add_files_state);
> > + status = remove_filename (notmuch, absolute, add_files_state);
> > talloc_free (absolute);
> > + if (status)
> > + return status;
> > }
> >
> > for (subdirs = notmuch_directory_get_child_directories (directory);
> > @@ -817,11 +822,14 @@ _remove_directory (void *ctx,
> > {
> > absolute = talloc_asprintf (ctx, "%s/%s", path,
> > notmuch_filenames_get (subdirs));
> > - _remove_directory (ctx, notmuch, absolute, add_files_state);
> > + status = _remove_directory (ctx, notmuch, absolute, add_files_state);
> > talloc_free (absolute);
> > + if (status)
> > + return status;
> > }
> >
> > notmuch_directory_destroy (directory);
> > + return NOTMUCH_STATUS_SUCCESS;
> > }
>
> In the two "return status" lines above seem to mean we don't call
> notmuch_directory_destroy. Does that matter?
Good point. I've fixed this to use the usual goto DONE cleanup idiom.
> The other query is not actually about this patch: just something that
> came up when reading it. Should notmuch_database_begin_atomic and
> notmuch_database_end_atomic always be paired? One of the (existing)
> error cases in remove_filename seems to return without calling end.
Yes, they should be. I've added a patch to fix that.
> Best wishes
>
> Mark
>
> > int
> > @@ -939,7 +947,9 @@ notmuch_new_command (void *ctx, int argc, char *argv[])
> >
> > gettimeofday (&tv_start, NULL);
> > for (f = add_files_state.removed_files->head; f && !interrupted; f = f->next) {
> > - remove_filename (notmuch, f->filename, &add_files_state);
> > + ret = remove_filename (notmuch, f->filename, &add_files_state);
> > + if (ret)
> > + goto DONE;
> > if (do_print_progress) {
> > do_print_progress = 0;
> > generic_print_progress ("Cleaned up", "messages",
> > @@ -950,7 +960,9 @@ notmuch_new_command (void *ctx, int argc, char *argv[])
> >
> > gettimeofday (&tv_start, NULL);
> > for (f = add_files_state.removed_directories->head, i = 0; f && !interrupted; f = f->next, i++) {
> > - _remove_directory (ctx, notmuch, f->filename, &add_files_state);
> > + ret = _remove_directory (ctx, notmuch, f->filename, &add_files_state);
> > + if (ret)
> > + goto DONE;
> > if (do_print_progress) {
> > do_print_progress = 0;
> > generic_print_progress ("Cleaned up", "directories",
next prev parent reply other threads:[~2012-04-22 4:21 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-27 15:49 [PATCH 0/3] Fix some error handling in notmuch new Austin Clements
2012-02-27 15:49 ` [PATCH 1/3] new: Consistently treat fatal errors as fatal Austin Clements
2012-04-16 15:53 ` Mark Walters
2012-04-22 4:11 ` Austin Clements
2012-02-27 15:49 ` [PATCH 2/3] new: Handle fatal errors in remove_filename and _remove_directory Austin Clements
2012-04-16 16:02 ` Mark Walters
2012-04-22 4:21 ` Austin Clements [this message]
2012-02-27 15:49 ` [PATCH 3/3] new: Print final fatal error message to stderr Austin Clements
2012-04-22 4:27 ` [PATCH v2 0/4] Fix some error handling in notmuch new Austin Clements
2012-04-22 4:27 ` [PATCH v2 1/4] new: Consistently treat fatal errors as fatal Austin Clements
2012-04-22 4:27 ` [PATCH v2 2/4] new: Handle fatal errors in remove_filename and _remove_directory Austin Clements
2012-04-22 7:34 ` Mark Walters
2012-04-22 15:36 ` Austin Clements
2012-04-22 4:27 ` [PATCH v2 3/4] new: Print final fatal error message to stderr Austin Clements
2012-04-22 4:27 ` [PATCH v2 4/4] new: Fix missing end_atomic in remove_filename on error Austin Clements
2012-04-22 7:42 ` Mark Walters
2012-04-22 15:50 ` [PATCH v3 0/4] Fix some error handling in notmuch new Austin Clements
2012-04-22 15:50 ` [PATCH v3 1/4] new: Consistently treat fatal errors as fatal Austin Clements
2012-04-25 2:28 ` David Bremner
2012-04-22 15:50 ` [PATCH v3 2/4] new: Handle fatal errors in remove_filename and _remove_directory Austin Clements
2012-04-22 15:50 ` [PATCH v3 3/4] new: Print final fatal error message to stderr Austin Clements
2012-04-22 15:50 ` [PATCH v3 4/4] new: Fix missing end_atomic in remove_filename on error Austin Clements
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
List information: https://notmuchmail.org/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20120422042105.GB17810@mit.edu \
--to=amdragon@mit.edu \
--cc=markwalters1009@gmail.com \
--cc=notmuch@notmuchmail.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.
Code repositories for project(s) associated with this public inbox
https://yhetil.org/notmuch.git/
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).