diff options
author | Eric Wong <e@yhbt.net> | 2020-06-10 07:04:29 +0000 |
---|---|---|
committer | Eric Wong <e@yhbt.net> | 2020-06-13 07:55:45 +0000 |
commit | c4dd2b484bbb20f2be0d7bc4a0ae0e91df635b4a (patch) | |
tree | df1069a4469010e1a1a8ca717a093357a563cdb7 /lib/PublicInbox/IMAPdeflate.pm | |
parent | afa44e1ddf1b9402caba5c7dc3c8e0f86194df86 (diff) | |
download | public-inbox-c4dd2b484bbb20f2be0d7bc4a0ae0e91df635b4a.tar.gz |
Since IMAP yields control to GitAsyncCat, IMAP->event_step may be invoked with {long_cb} still active. We must be sure to bail out of IMAP->event_step if that happens and continue to let GitAsyncCat drive IMAP. This also improves fairness by never processing more than one request per ->event_step.
Diffstat (limited to 'lib/PublicInbox/IMAPdeflate.pm')
-rw-r--r-- | lib/PublicInbox/IMAPdeflate.pm | 10 |
1 files changed, 10 insertions, 0 deletions
diff --git a/lib/PublicInbox/IMAPdeflate.pm b/lib/PublicInbox/IMAPdeflate.pm index 67c9a973..42daa6cf 100644 --- a/lib/PublicInbox/IMAPdeflate.pm +++ b/lib/PublicInbox/IMAPdeflate.pm @@ -59,6 +59,16 @@ sub do_read ($$$$) { $doff = length($dbuf); my $r = PublicInbox::DS::do_read($self, \$dbuf, $len, $doff) or return; + # Workaround inflate bug appending to OOK scalars: + # <https://rt.cpan.org/Ticket/Display.html?id=132734> + # We only have $off if the client is pipelining, and pipelining + # is where our substr() OOK optimization in event_step makes sense. + if ($off) { + my $copy = $$rbuf; + undef $$rbuf; + $$rbuf = $copy; + } + # assert(length($$rbuf) == $off) as far as NNTP.pm is concerned # -ConsumeInput is true, so $dbuf is automatically emptied my $err = $zin->inflate($dbuf, $rbuf); |