Eric Wong [Fri, 24 Jul 2020 05:56:05 +0000 (05:56 +0000)]
searchidx: $batch_cb => v1_checkpoint
Another closure gone, and we may be able to share more
code with v2 in upcoming commits.
Eric Wong [Fri, 24 Jul 2020 05:56:04 +0000 (05:56 +0000)]
searchidx: support async git check
This allows v1 indexing to run while the `cat-file --batch-check'
process is waiting on high-latency storage.
Eric Wong [Fri, 24 Jul 2020 05:56:03 +0000 (05:56 +0000)]
v2writable: share log2stack code with v1
Another step in making v1 and v2 more similar.
Eric Wong [Fri, 24 Jul 2020 05:56:02 +0000 (05:56 +0000)]
index+xcpdb: support --no-sync flag
This allows us to speed up indexing operations to SQLite
and Xapian.
Unfortunately, it doesn't affect operations using
`xapian-compact' and the compactor API, since that doesn't seem
to support Xapian::DB_NO_SYNC, yet.
Eric Wong [Fri, 24 Jul 2020 05:56:01 +0000 (05:56 +0000)]
searchidx: make v1 indexing closer to v2
We'll switch to using IdxStack here to ensure we get repeatable
results and ascending THREADIDs according to git chronology.
This means we'll need a two-pass reindex to index existing
messages before indexing new messages.
Since we no longer have a long-lived git-log process, we don't
have to worry about old Xapian referencing the git-log pipe
w/o FD_CLOEXEC, either.
Eric Wong [Fri, 24 Jul 2020 05:56:00 +0000 (05:56 +0000)]
searchidx: rename _xdb_{acquire,release} => idx_
The "xdb" prefix was inaccurate since it's used by
indexlevel=basic, which is Xapian-free. The '_' (underscore)
prefix was also wrong for a method which is called across
package boundaries.
Eric Wong [Fri, 24 Jul 2020 05:55:59 +0000 (05:55 +0000)]
xapcmd: set {from} properly for v1 inboxes
This was a bug, but I'm not sure where it matters, yet, but it
may matter in the future.
Eric Wong [Fri, 24 Jul 2020 05:55:58 +0000 (05:55 +0000)]
v2writable: clarify "epoch" comment
Eric Wong [Fri, 24 Jul 2020 05:55:57 +0000 (05:55 +0000)]
v2writable: get rid of {reindex_pipe} field
Since normal per-epoch indexing no longer holds a "git log"
process open, we don't need to worry about not sharing the
pipe with forked shards when we restart the indexer.
While we're in the area, better describe what `unindex' does,
since it's a rarely-used but necessary code path.
Eric Wong [Fri, 24 Jul 2020 05:55:56 +0000 (05:55 +0000)]
v2writable: use read-only PublicInbox::Git for cat_file
We can reduce the number of parameters we pass around on stack
and make our read-write and read-only code paths more uniform.
Eric Wong [Fri, 24 Jul 2020 05:55:55 +0000 (05:55 +0000)]
search: avoid copying {inboxdir}
Instead, storing {xdir} will allow us to avoid string
concatenation in the read-only path and save us a little
hash entry space.
Eric Wong [Fri, 24 Jul 2020 05:55:54 +0000 (05:55 +0000)]
use consistent {ibx} field for writable code paths
This is a step which makes our use of abbreviations more
consistent when referring to PublicInbox::Inbox objects.
We'll also be reducing the number of redundant fields
in SearchIdx and V2Writable code paths to make the
object graph easier-to-follow.
Eric Wong [Fri, 24 Jul 2020 05:55:53 +0000 (05:55 +0000)]
v2writable: drop "EPOCH.git indexing $RANGE" progress
It'll be one continuous range with IdxStack.
Eric Wong [Fri, 24 Jul 2020 05:55:52 +0000 (05:55 +0000)]
v2writable: allow >= 40 byte git object IDs
Another step in slowly updating our code to support SHA-256 or
whatever other hash algorithms git may support in the future.
Eric Wong [Fri, 24 Jul 2020 05:55:51 +0000 (05:55 +0000)]
v2writable: move {autime} and {cotime} into $sync state
The V2Writable object may be long-lived, so it makes more
sense to put the {autime} and {cotime} fields into the
shorter-lived index_sync state.
Eric Wong [Fri, 24 Jul 2020 05:55:50 +0000 (05:55 +0000)]
v2writable: index_sync: reduce fill_alternates calls
Instead of doing fill_alternates for every epoch we're indexing,
just do it once at the start of index_sync invocation. This
will set us up for using a single "git cat-file" process for
indexing multiple epochs.
Eric Wong [Fri, 24 Jul 2020 05:55:49 +0000 (05:55 +0000)]
v2writable: introduce idx_stack
This avoids pinning a potentially large chunk of memory from
`git-log --reverse' into RAM (or triggering less predictable
swap behavior). Instead it uses a contiguous temporary file
with a fixed-size record for every blob we'll need to index.
Eric Wong [Fri, 24 Jul 2020 05:55:48 +0000 (05:55 +0000)]
v2: index forwards (via `git log --reverse')
Since we'll need to expose THREADID to JMAP and IMAP users,
index all messages in the order they were committed to ensure
our `tid' (thread ID) column ascends in mirrors the same way
they do in the source inbox.
This drastically simplifies our code but increases memory
usage of `git-log'. The next commit will bring memory use
back down at the expense of $TMPDIR usage.
Eric Wong [Fri, 24 Jul 2020 05:55:47 +0000 (05:55 +0000)]
index: support --rethread switch to fix old indices
Older versions of public-inbox < 1.3.0 had subtly
different semantics around threading in some corner
cases. This switch (when combined with --reindex)
allows us to fix them by regenerating associations.
Eric Wong [Sat, 18 Jul 2020 06:36:32 +0000 (06:36 +0000)]
msgmap: fix atfork_* callbacks
Noticed while reindexing a largish v2 inbox in parallel on an
SSD which required checkpointing and respawning shard workers.
Fixes: f06e84220e5566e7 ("over+msgmap: do not store filename after DBI->connect")
Eric Wong [Fri, 17 Jul 2020 07:25:07 +0000 (07:25 +0000)]
v2writable: git_hash_raw: avoid $TMPDIR write
We can rely on FD_CLOEXEC being set by default (since Perl 5.6+)
on pipes to avoid FS/page-cache traffic, here. We also know
"git hash-object" won't output anything until it's consumed all
of its standard input; so there's no danger of a deadlock even
in the the unlikely case git uses a hash that can't fit into
PIPE_BUF :P
Eric Wong [Fri, 17 Jul 2020 06:31:55 +0000 (06:31 +0000)]
search: simplify unindexing
Since over.sqlite3 seems here to stay, we no longer need to do
Message-ID lookups against Xapian and can simply rely on the
docid <=> NNTP article number equivalancy SCHEMA_VERSION=15
gave us.
This rids us of the closure-using batch_do sub in the v1
code path and vastly simplifies both v1 and v2 unindexing.
Eric Wong [Fri, 17 Jul 2020 06:31:54 +0000 (06:31 +0000)]
searchidx: use v5.10.1, parent.pm, drop warnings
Prefer "parent" to "base" since the former is lighter and part
of Perl 5.10+. We'll also rely on warnings from "-w" globally
(or not) instead of via "use".
Eric Wong [Fri, 17 Jul 2020 06:31:53 +0000 (06:31 +0000)]
overidx: favor non-OO sub dispatch for internal subs
OO method dispatch was 10-15% slower when I was implementing the
NNTP server. It also serves as a helpful reminder to the reader
at the callsite as to whether a sub is likely in the same
package as the caller or not.
Eric Wong [Fri, 17 Jul 2020 06:31:52 +0000 (06:31 +0000)]
overidx: each_by_mid: pass self and args to callbacks
This saves runtime allocations and reduces the likelyhood of
memory leaks either from cycles or buggy old Perl versions.
Eric Wong [Fri, 17 Jul 2020 06:31:51 +0000 (06:31 +0000)]
with_umask: pass args to callback
While it makes the code flow slightly less well in some places,
it saves us runtime allocations and indentation.
Eric Wong [Fri, 17 Jul 2020 06:31:50 +0000 (06:31 +0000)]
import: use common capitalization for filtering headers
In case this ends up in the same process as Mbox::msg_hdr,
it can reduce memory use by sharing the cache key in
PublicInbox::Eml::re_memo
Eric Wong [Fri, 17 Jul 2020 06:31:49 +0000 (06:31 +0000)]
drop binmode usage
We only support Unix-like platforms where binmode (":raw") is
the default anyways, and v5.10 semantics means it won't do
unicode_strings (unlike v5.12). So save some lines of code.
Eric Wong [Fri, 17 Jul 2020 06:31:48 +0000 (06:31 +0000)]
v2: use v5.10.1, parent.pm, drop warnings
The "5.010_001" form was for Perl 5.6, which I doubt anybody
would attempt; so favor "v5.10.1" as it is more readable to
humans. Prefer "parent" to "base" since the former is lighter.
We'll also rely on warnings from "-w" globally (or not) instead
of via "use".
We'll also update "use" statements to reflect what's actually
used by V2Writable.
Eric Wong [Thu, 16 Jul 2020 21:54:24 +0000 (21:54 +0000)]
t/import: quiet warning, clobber variable
The eval in key2sub via t/run.perl ("make check-run") won't
trigger the warning, but running "prove -bvw t/import.t"
directly, does. In any case, ensure the contents of this
variable doesn't linger across runs.
Eric Wong [Fri, 17 Jul 2020 03:57:24 +0000 (03:57 +0000)]
doc: add some recommendations around slow HDDs
grok-pull is still painful with serialization on an old USB 2.0
HDD, but at least it can finish with flock(1) and disabling
parallelization. While parallel "git fetch" doesn't seem so
bad, slow seeks are exacerbated by parallel reads in Xapian.
That means some updates can take days instead of hours. The
same updates take only seconds or minutes on an SSD.
Eric Wong [Fri, 17 Jul 2020 06:07:10 +0000 (06:07 +0000)]
config: reject `\n' in `inboxdir'
"\n" and other characters requiring quoting and/or escaping in
in $GIT_DIR/objects/info/alternates was not supported in git 2.11
and earlier; nor does it seem supported at all in libgit2.
This will allow us to support sharing git-cat-file or similar
endpoints across multiple inboxes via alternates.
This breaks an existing use case for anybody wacky
enough to put `\n' in the `inboxdir' pathname; but I doubt
this affects anybody.
Eric Wong [Tue, 14 Jul 2020 10:06:40 +0000 (10:06 +0000)]
doc: release notes and version info updates
Update release notes with some features in the 1.6 timeline.
We'll note the version availability of some command-line
options, it may help users who are reading the latest
documentation online but running older versions.
Eric Wong [Tue, 14 Jul 2020 02:14:32 +0000 (02:14 +0000)]
over+msgmap: do not store filename after DBI->connect
SQLite already knows the filename internally, so avoid having it
as a long-lived Perl SV to save some bytes when there's many
inboxes and open DBs.
Eric Wong [Tue, 14 Jul 2020 02:14:31 +0000 (02:14 +0000)]
nntpd+imapd: detect unlinked msgmap
While it's even less common to experience a replaced
msgmap.sqlite3 file, BOFHs may do the darndest things. This is
another step towards reducing the number of needless wakeups
we need to do in long-lived read-only daemons.
Eric Wong [Tue, 14 Jul 2020 02:14:30 +0000 (02:14 +0000)]
over: unset sqlite_unicode attribute
None of the human-readable strings stored in over.sqlite3
require UTF-8. Message-IDs do not, nor do the compressed
Subject IDs (sid) we use for Subject-based threading. And the
`ddd' (doc-data-deflated) column is of course binary data.
This frees us of having to use SQL_BLOB for the `ddd', column,
and will open the door for us to use dbh_new for Msgmap, too.
Eric Wong [Sun, 5 Jul 2020 21:51:31 +0000 (21:51 +0000)]
xapcmd: delay over->check_inodes trigger
We must not trigger wakeups on InboxIdle users until after we've
renamed all files into place. Otherwise, the InboxIdle caller
may just reopen the old (soon-to-be-unlinked) file.
This fixes occasional test failures in t/nntpd.t
Fixes: f977826a17f8735e ("lock: reduce inotify wakeups")
Eric Wong [Mon, 13 Jul 2020 02:23:28 +0000 (02:23 +0000)]
xt/mem-imapd-tls: avoid EMFILE in -imapd process
Test::More dups standard FDs and may create FDs for other
purposes. run_mode => 0 lets us rely on FD_CLOEXEC to ensure
-imapd has enough FDs to accept all incoming connections at
the cost of higher (one-off) startup time.
Eric Wong [Mon, 13 Jul 2020 07:00:35 +0000 (07:00 +0000)]
imap: SEARCH fails more gracefully in non-slice mailbox
Instead of returning "BAD program fault", just give the
standard "BAD search not available"... message we show
for mailbox slices.
Eric Wong [Tue, 7 Jul 2020 20:37:38 +0000 (20:37 +0000)]
wwwatomstream: avoid uninitialized warnings for $email
As in Import, we'll fall back to Sender: if From: is missing,
and use the primary_address of the inboxes to indicate the total
absence of those fields.
Eric Wong [Tue, 7 Jul 2020 20:37:37 +0000 (20:37 +0000)]
imap: avoid warnings on non-slice mailboxes
Non-slice mailboxes never have messages themselves,
so we must not assume a message exists when sending
untagged EXISTS messages.
Eric Wong [Tue, 7 Jul 2020 20:37:36 +0000 (20:37 +0000)]
viewvcs: stop checking unused "B" query parameter
The resulting OID ("oid_b") is a required arg and part of
$env->{PATH_INFO}, instead; so it's never part of an optional
query parameter.
Eric Wong [Tue, 7 Jul 2020 20:37:35 +0000 (20:37 +0000)]
hval: to_filename: return `undef' instead of empty string
Returning an empty string for a filename makes no sense,
so instead return `undef' so the caller can setup a fallback
using the "//" operator.
This fixes uninitialized variable warnings because split()
on an empty string returns `undef', which caused to_filename
to warn on s// and tr// ops.
Eric Wong [Tue, 7 Jul 2020 20:37:34 +0000 (20:37 +0000)]
viewvcs: allow "0" as a path name
This means we need to filter out "" from query parameters.
While we're at it, update comments for the WWW endpoint.
Eric Wong [Wed, 8 Jul 2020 20:58:49 +0000 (20:58 +0000)]
doc: standards: link IMAP capabilities and response codes
We'll be implementing some IMAP search/threading extensions in
IMAP and providing analogues over HTTP via JMAP.
Eric Wong [Tue, 7 Jul 2020 06:17:24 +0000 (06:17 +0000)]
t/spawn: fix test reliability
Since Perl doesn't internally use a self-pipe for
sleep/select/poll/etc, wake up every 10ms to ensure
it can see the SIGCHLD; since neither signalfd nor EVFILT_SIGNAL
are always available.
Fixes: 761baa2a300e4268 ("spawn: unblock SIGCHLD in subprocess")
Eric Wong [Fri, 19 Jun 2020 17:18:47 +0000 (17:18 +0000)]
mda: use PublicInbox::Address to extract emails
I'm not sure why this wasn't done in Jun/July 2016 when I was
working on PublicInbox::Address to replace the DoS-vulnerable
Email::Address.
Nowadays, PublicInbox::Address allows using Email::Address::XS
which should be fast and robust.
Eric Wong [Sun, 5 Jul 2020 22:44:52 +0000 (22:44 +0000)]
doc/technical/whyperl: note Perl 7 announcement
Right now[1] the Perl upstream plan is to maintain 5 compatibility
in Perl 7 for at least 5 years[1], and perhaps drop it when Perl 8
comes along. That said, distros may pick it and maintain 5 on their
own given the vast amounts of perfectly good legacy code out there.
[1] http://nntp.perl.org/group/perl.perl5.porters/257817
[2] http://nntp.perl.org/group/perl.perl5.porters/257565
Eric Wong [Sun, 5 Jul 2020 22:44:51 +0000 (22:44 +0000)]
doc/technical/whyperl: reword bit around installed docs
I originally proposed this rewording to address Leah's comment
but forgot to squash it in :x
Link: https://public-inbox.org/meta/20200408221741.GA10142@dcvr/
Cc: Leah Neukirchen <leah@vuxu.org>
Eric Wong [Sun, 5 Jul 2020 22:44:50 +0000 (22:44 +0000)]
doc: daemon: update documentation around Inline::C
`~/.cache/public-inbox/inline-c' is supported, nowadays
for convenience, but Inline::C usage will remain opt-in.
Eric Wong [Sun, 5 Jul 2020 22:44:49 +0000 (22:44 +0000)]
git: use v5.10.1, parent.pm and Time::HiRes::stat
parent.pm is leaner than base.pm, and Time::HiRes::stat is
more accurate, so take advantage of these Perl 5.10+-isms
since it's been over a year since we left 5.8 behind.
Eric Wong [Sun, 5 Jul 2020 23:27:59 +0000 (23:27 +0000)]
gzipfilter: check http->{forward} for client disconnects
We actually don't do anything with {env} or {'psgix.io'}
on client aborts, so checking the truthiness of '{forward}'
is necessary.
Eric Wong [Sun, 5 Jul 2020 23:27:58 +0000 (23:27 +0000)]
daemon: warn on missing blobs
Since -edit and -purge should be rare and TOCTOU around them
rarer still; missing {blobs} could be indicative of a real bug
elsewhere. Warn on them.
And I somehow ended up with 3 different field names for Inbox
objects. Perhaps they'll be made consistent in the future.
Eric Wong [Sun, 5 Jul 2020 23:27:57 +0000 (23:27 +0000)]
gzipfilter: drop HTTP connection on bugs or data corruption
While all the {async_next} callbacks needed eval guards anyways
because of DS->write, {async_eml} callbacks did not.
Ensure any bugs in our code or data corruption result in
termination of the HTTP connection, so as not to leave clients
hanging on a response which never comes or is mangled in some
way.
Eric Wong [Sun, 5 Jul 2020 23:27:56 +0000 (23:27 +0000)]
wwwattach: support async blob retrievals
We can reuse some of the GzipFilter infrastructure used by other
WWW components to handle slow blob retrieval, here. The
difference from previous changes is we don't decide on the 200
status code until we've retrieved the blob and found the
attachment.
While we're at it, ensure we can compress text attachment
responses once again, since all text attachments are served
as text/plain.
Eric Wong [Sun, 5 Jul 2020 23:27:55 +0000 (23:27 +0000)]
wwwtext: simplify gzf_maybe use
gzf_maybe always returns a GzipFilter object, even if it uses
CompressNoop. We can also use ->zflush instead of
->translate(undef) here for the final bit.
Eric Wong [Sun, 5 Jul 2020 23:27:54 +0000 (23:27 +0000)]
view: simplify eml_entry callers further
This simplifies the primary callers of eml_entry while only making
mknews.perl worse.
Eric Wong [Sun, 5 Jul 2020 23:27:53 +0000 (23:27 +0000)]
www: update internal docs
We no longer favor getline+close for streaming PSGI responses
when using public-inbox-httpd. We still support it for other
PSGI servers, though.
Eric Wong [Sun, 5 Jul 2020 23:27:52 +0000 (23:27 +0000)]
wwwstream: eliminate ::response, use html_oneshot
All of our streaming responses use ::aresponse, now, and our
synchronous responses use html_oneshot. So there's no need
for the old WwwStream::response.
Eric Wong [Sun, 5 Jul 2020 23:27:51 +0000 (23:27 +0000)]
view: /$INBOX/$MSGID/t/: avoid extra hash lookup in eml case
We can build and buffer the HTML <head> section once the first
non-ghost message in a thread is loaded, so there's no need to
perform an extra check on $ctx->{nr} once the $eml is ready.
Eric Wong [Sun, 5 Jul 2020 23:27:50 +0000 (23:27 +0000)]
view: eml_entry: reduce parameters
We can save stack space and simplify subroutine calls, here.
Eric Wong [Sun, 5 Jul 2020 23:27:49 +0000 (23:27 +0000)]
ssearchview: /$INBOX/?q=$QUERY&x=t uses async blobs
Another 10% or so speedup when displaying full messages off
search results.
Eric Wong [Sun, 5 Jul 2020 23:27:48 +0000 (23:27 +0000)]
feed: /$INBOX/new.html fetches blobs asynchronously
Once again this speeds another endpoint up 10% or so.
Eric Wong [Sun, 5 Jul 2020 23:27:47 +0000 (23:27 +0000)]
feed: generate_i: eliminate pointless loop
$ctx->{msgs} won't ever contain undef values.
Eric Wong [Sun, 5 Jul 2020 23:27:46 +0000 (23:27 +0000)]
view: update /$INBOX/$MSGID/T/ to be async
Another 10% or so speedup in a frequently-hit endpoint.
Eric Wong [Sun, 5 Jul 2020 23:27:45 +0000 (23:27 +0000)]
view: /$INBOX/$MSGID/t/ reads blobs asynchronously
Once again, this shows a ~10% speedup with multi-message
threads in xt/httpd-async-stream.t regardless of whether
TEST_JOBS is 1 or 100.
Eric Wong [Sun, 5 Jul 2020 23:27:44 +0000 (23:27 +0000)]
view: make /$INBOX/$MSGID/ permalink async
This will allow -httpd to handle other requusts if waiting on
an HDD seek or git to decode a blob.
Eric Wong [Sun, 5 Jul 2020 23:27:43 +0000 (23:27 +0000)]
wwwstream: subclass off GzipFilter
This makes WwwStream closer to MboxGz and WwwAtomStream
and will eventually allow us to follow the same patterns.
Eric Wong [Sun, 5 Jul 2020 23:27:42 +0000 (23:27 +0000)]
wwwstream: use parent.pm and no warnings
parent.pm is leaner than base and we'll rely on `-w' for
warnings during development.
Eric Wong [Sun, 5 Jul 2020 23:27:41 +0000 (23:27 +0000)]
remove unused/redundant zlib-related imports
Z_FINISH is the default for Compress::Raw::Zlib::Deflate->flush,
anyways, so there's no reason to import it. And none of C::R::Z
is needed in WwwText now that gzf_maybe handles it all.
Eric Wong [Sun, 5 Jul 2020 23:27:40 +0000 (23:27 +0000)]
www: start making gzipfilter the parent response class
Virtually all of our responses are going to be gzipped, anyways.
This will allow us to utilize zlib as a buffering layer and
share common code for async blob retrieval responses.
To streamline this and allow GzipFilter to be a parent class,
we'll replace the NoopFilter with a similar CompressNoop class
which emulates the two Compress::Raw::Zlib::Deflate methods we
use.
This drops a bunch of redundant code and will hopefully make
upcoming WwwStream changes easier to reason about.
Eric Wong [Sun, 5 Jul 2020 23:27:39 +0000 (23:27 +0000)]
wwwstream: reduce blob fetch paths for ->getline
This will make it easier to support asynchronous blob
retrievals. The `$ctx->{nr}' counter is no longer implicitly
supplied since many users didn't care for it, so stack overhead
is slightly reduced.
Eric Wong [Sun, 5 Jul 2020 23:27:38 +0000 (23:27 +0000)]
wwwstream: reduce object graph depth
Like with WwwAtomStream and MboxGz, we can bless the existing
$ctx object directly to avoid allocating a new hashref. We'll
also switch from "->" to "::" to reduce stack utilization.
Eric Wong [Sun, 5 Jul 2020 23:27:37 +0000 (23:27 +0000)]
wwwatomstream: support async blob fetch
This allows -httpd to handle other requests while waiting
for git to retrieve and decode blobs. We'll also break
apart t/psgi_v2.t further to ensure tests run against
-httpd in addition to generic PSGI testing.
Using xt/httpd-async-stream.t to test against clones of meta@public-inbox.org
shows a 10-12% performance improvement with the following env:
TEST_JOBS=1000 TEST_CURL_OPT=--compressed TEST_ENDPOINT=new.atom
Eric Wong [Sun, 5 Jul 2020 23:27:36 +0000 (23:27 +0000)]
xt/httpd-async-stream: allow more options
We want to be able to parallelize and stress test more
endpoints and toggle `--compressed' and possibly other
options in curl.
Eric Wong [Sun, 5 Jul 2020 23:27:35 +0000 (23:27 +0000)]
wwwatomstream: reuse $ctx as $self
No need to deepen our object graph, here.
Eric Wong [Sun, 5 Jul 2020 23:27:34 +0000 (23:27 +0000)]
wwwatomstream: use PublicInbox::Inbox->modified for feed_updated
stat(2) on the inboxdir is unlikely to be correct, now that
msgmap truncates its journal (rather than unlinking it).
Eric Wong [Sun, 5 Jul 2020 23:27:33 +0000 (23:27 +0000)]
wwwatomstream: simplify feed_update callers
We always return Z (UTC) times, anyways, so we'll always
use gmtime() on the seconds-after-the-epoch.
Eric Wong [Sun, 5 Jul 2020 23:27:32 +0000 (23:27 +0000)]
mbox: async blob fetch for "single message" raw mboxrd
This restores gzip-by-default behavior for /$INBOX/$MSGID/raw
endpoints for all indexed inboxes. Unindexed v1 inboxes will
remain uncompressed, for now.
Eric Wong [Sun, 5 Jul 2020 23:27:31 +0000 (23:27 +0000)]
mboxgz: reduce hash depth
We can bless $ctx directly into a MboxGz object to reduce
hash lookups and allocations.
Eric Wong [Sun, 5 Jul 2020 23:27:30 +0000 (23:27 +0000)]
mboxgz: do asynchronous git blob retrievals
This lets the -httpd worker process make better use of time
instead of waiting for git-cat-file to respond. With 4 jobs in
the new test case against a clone of
<https://public-inbox.org/meta/>, a speedup of 10-12% is shown.
Even a single job shows a 2-5% improvement on an SSD.
Eric Wong [Sun, 5 Jul 2020 23:27:29 +0000 (23:27 +0000)]
stop auto-loading Plack::Middleware::Deflater
Instead of gzipping some (mbox.gz, manifest.js.gz) responses and
leaving P::M::D to do the rest, we gzip everything ourselves,
now, so P::M::D is redundant.
Eric Wong [Sun, 5 Jul 2020 23:27:28 +0000 (23:27 +0000)]
qspawn: learn to gzip streaming responses
This will allow us to gzip responses generated by cgit
and any other CGI programs or long-lived streaming
responses we may spawn.
Eric Wong [Sun, 5 Jul 2020 23:27:27 +0000 (23:27 +0000)]
wwwstatic: support gzipped directory listings
This will allow others to mimic our award-winning homepage
design without needing to rely on Plack::Middleware::Deflater
or varnish to compress responses.
Eric Wong [Sun, 5 Jul 2020 23:27:26 +0000 (23:27 +0000)]
mbox: remove html_oneshot import
It's no longer needed, we no longer show a runtime error
for zlib being missing, as zlib is a hard requirement.
Fixes: a318e758129d616b ("make zlib-related modules a hard dependency")
Eric Wong [Sun, 5 Jul 2020 23:27:25 +0000 (23:27 +0000)]
{gzip,noop}filter: ->zmore returns undef, always
This simplifies callers, as witnessed by the change to
WwwListing. It adds overhead to NoopFilter, but NoopFilter
should see little use as nearly all HTTP clients request gzip.
Eric Wong [Sun, 5 Jul 2020 23:27:24 +0000 (23:27 +0000)]
gzipfilter: replace Compress::Raw::Deflate usages
The new ->zmore and ->zflush APIs make it possible to replace
existing verbose usages of Compress::Raw::Deflate and simplify
buffering logic for streaming large gzipped data.
One potentially user visible change is we now break the mbox.gz
response on zlib failures, instead of silently continuing onto
the next message. zlib only seems to fail on OOM, which should
be rare; so it's ideal we drop the connection anyways.
Eric Wong [Sun, 5 Jul 2020 23:27:23 +0000 (23:27 +0000)]
wwwlisting: use GzipFilter for HTML
The changes to GzipFilter here may be beneficial for building
HTML and XML responses in other places, too.
Eric Wong [Sun, 5 Jul 2020 23:27:22 +0000 (23:27 +0000)]
www: need: use WwwStream::html_oneshot
It'll give us a nicer HTML header and footer.
Eric Wong [Sun, 5 Jul 2020 23:27:21 +0000 (23:27 +0000)]
wwwtext: switch to html_oneshot
No point in streaming a tiny response via ->getline,
but we may stream to a gzipped buffer, later.
Eric Wong [Sun, 5 Jul 2020 23:27:20 +0000 (23:27 +0000)]
wwwtext: gzip text/plain responses, as well
Most of our plain-text responses are config files
big enough to warrant compression.
Eric Wong [Sun, 5 Jul 2020 23:27:19 +0000 (23:27 +0000)]
www*stream: gzip ->getline responses
Our most common endpoints deserve to be gzipped.
Eric Wong [Sun, 5 Jul 2020 23:27:18 +0000 (23:27 +0000)]
wwwstream: oneshot: perform gzip without middleware
Plack::Middleware::Deflater forces us to use a memory-intensive
closure. Instead, work towards building compressed strings in
memory to reduce the overhead of buffering large HTML output.
Eric Wong [Sun, 5 Jul 2020 23:27:17 +0000 (23:27 +0000)]
gzipfilter: minor cleanups
We currently don't use bytes::length in ->write, so there's no
need to `use bytes'. Favor `//=' to describe the intent of the
conditional assignment since the C::R::Z::Deflate object is
always truthy. Also use the local $gz variable to avoid
unnecessary {gz} hash lookups.
Eric Wong [Mon, 6 Jul 2020 06:11:02 +0000 (06:11 +0000)]
git_async_cat: unref pipes on EOF from git->cleanup
We avoided a managed circular reference in
10ee3548084c125f
but introduced a pipe FD leak, instead. So handle the EOF
we get when the "git cat-file --batch" process exits and
closes its stdout FD.
v2: remove ->close entirely. PublicInbox::Git->cleanup
handles all cleanup. This prevents us from inadvertantly
deleting the {async_cat} field associated with a different
pipe than the one GAC is monitoring.
Fixes: 10ee3548084c125f ("git_async_cat: remove circular reference")
Eric Wong [Mon, 6 Jul 2020 04:20:12 +0000 (04:20 +0000)]
inboxidle: warn on inotify_add_watch failures
inotify_add_watch(2), open(2), stat(2) may all fail due to
permissions errors, especially when running -nntpd/-imapd
as `nobody' as recommended.
Eric Wong [Sat, 4 Jul 2020 20:25:25 +0000 (20:25 +0000)]
t/import: test for nasty characters
Spammers may send emails with nasty characters which can throw
off git-fast-import. Users with non-existent or weaker spam
filters may be susceptible to corruption in the fast-import
stream as a result.
This was actually quietly fixed in git on 2020-06-01 by
commit
9ab886546cc89f37819e1ef09cb49fd9325b3a41
("smsg: introduce ->populate method"), but no test case
was created.
Reported-by: Eric W. Biederman <ebiederm@xmission.com>
Link: https://public-inbox.org/meta/87imf4qn87.fsf@x220.int.ebiederm.org/
Link: https://public-inbox.org/meta/20200601100657.14700-6-e@yhbt.net/
Eric Wong [Sat, 4 Jul 2020 21:33:13 +0000 (21:33 +0000)]
watch: don't burn CPU on IDLE failures
Network connections fail and need to be detected sooner rather
than later during IDLE to avoid backtrace floods. In case the
IDLE process dies completely, don't respawn right away, either,
to avoid entering a respawn loop.
There's also a typo fix :P
Eric Wong [Thu, 2 Jul 2020 20:12:40 +0000 (20:12 +0000)]
spawn: drop unused sys/uio.h include
We no longer use writev(2) in pi_fork_exec to emit errors.
Eric Wong [Wed, 1 Jul 2020 21:06:19 +0000 (21:06 +0000)]
overidx: document why we don't use SQLite WAL
I was wondering about this myself the other day and had to read
up on it. So make a note of it for future readers.