This is needed to support forking from already-forked lei workers
and $lei->{2} is already STDERR.
Fixes: e015c3742f91 (lei: use autodie where appropriate, 2023-10-17)
---
lib/PublicInbox/LEI.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/PublicInbox/LEI.pm
We should be able to treat v2 outputs just like any other mail
format, with the exception that content dedupe is always
enforced by the v2 format.
This allows users hosting v2 public-inboxes to catch up broken
synchronization from alternate archives such as the mbox
archives hosted by
Eric Wong wrote:
> Avoid mixing autodie use in different scopes since it's likely
> to cause problems like it did in Gcf2. While none of these
> fix known problems with test cases, it's likely worthwhile to
> avoid it anyways to avoid future surprises.
> lib/PublicInbox/XapHe
Eric Wong wrote:
> +++ b/t/cindex.t
> @@ -210,7 +210,7 @@ EOM
> my $cmd = [ qw(-cindex -u --all --associate -d), "$tmp/ext",
> '-I', $basic->{inboxdir} ];
> $cidx_out = $cidx_err = '';
> - ok(run_script($cmd, $env, $opt), 'ass
These only spring up with libgit2-devel installed, but
3/3 seems important for all users of older Perl in
order to avoid future surprises.
Eric Wong (3):
gcf2client: add alias for PublicInbox::Git::fail
gcf2: fix autodie usage for older Perl
treewide: more autodie safety fixes for older
Ensure we can ->fail properly from other subs we can within
Gcf2Client. This doesn't fix the test failures on CentOS 7.x,
but tries to make it easier to fix underlying problems and
report OOM errors and other things which the test suite doesn't
touch on.
---
lib/PublicInbox/Gcf2Client.pm | 1 +
Avoid mixing autodie use in different scopes since it's likely
to cause problems like it did in Gcf2. While none of these
fix known problems with test cases, it's likely worthwhile to
avoid it anyways to avoid future surprises.
For Process::IO, we'll add some additional tests in t/io.t
to ensure
At least on Perl v5.16.3 on CentOS 7.x, use-ing autodie within
BEGIN {} affects all subroutines in that package, too. So just
use autodie at the top-level and rely on CORE::* and try_cat
to handle cases where autodie isn't desired.
---
lib/PublicInbox/Gcf2.pm | 13 ++---
1 file changed,
Konstantin Ryabitsev wrote:
> t/gcf2_client.t .. 1/? Can't locate object method "fail" via
> package "PublicInbox::Gcf2Client" at
> /home/mricon/public-inbox-test/blib/lib/PublicInbox/Git.pm line 269.
> (in cleanup) Can't locate object method "fail" via package
>
Konstantin Ryabitsev wrote:
> There are two sources of potential discrepancy:
>
> - differences in CPAN module versions I have installed
> - different git or xapian14 versions
Could also be CPU count or CPU speeds since IPC bugs are
sensitive to that. That said, I can't get cindex.t to fail
Konstantin Ryabitsev wrote:
> Looks like the last time I am able to successfully run "make test" is before
> this commit:
>
> b231d91f42d791becf7b6861e723833d71e73237 is the first bad commit
>
> The error I start getting after this commit is:
>
> t/extsearch.t 160/?
> #
Unlike modern Perls, Perl 5.16.3 on CentOS doesn't accept
negative string signals like "-TERM" .
This only became a problem since commit b231d91f42d7
(treewide: enable warnings in all exec-ed processes)
made our code stricter by enabling more warnings.
In both cases, the kill is probably
Neither of these fix the t/cindex.t stuck problem Konstantin is
encountering, though...
Eric Wong (2):
lei: use -signal numbers for old Perl
t/lei-import: account for more verbose error
lib/PublicInbox/LEI.pm| 2 +-
lib/PublicInbox/LeiXSearch.pm | 2 +-
t/lei-import.t
Perl 5.16.3 on CentOS seems more verbose in one of the EIO
tests. Relax the regexp so we can account for extra errors
reported by Perl.
---
t/lei-import.t | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/t/lei-import.t b/t/lei-import.t
index bd562617..b4446b56 100644
---
Konstantin Ryabitsev wrote:
> On Tue, Nov 14, 2023 at 10:46:57PM +0000, Eric Wong wrote:
> > > I can't do +E because that's not available to me under CentOS7 (I can't
> > > wait
> > > until we move on, but just when we think the yak is fully shaved, we find
> &
Konstantin Ryabitsev wrote:
> On Tue, Nov 14, 2023 at 10:16:53PM +0000, Eric Wong wrote:
> > Konstantin Ryabitsev wrote:
> > > └─-cindex -u --al,4432
> > > ├─cidx shard[0],4646
> > > └─cidx shard[1],4647
> > &g
Konstantin Ryabitsev wrote:
> └─-cindex -u --al,4432
> ├─cidx shard[0],4646
> └─cidx shard[1],4647
>
> Anything I can do to figure out why this is happening?
You can show me strace and lsof +E of the processes (any other
processes (join|sort|awk|perl)?). This
Thanks, pushed as commit 58e6ee9df4f74b1078541c8924cf2918ceec0765
Noticed while working on another feature...
---
lib/PublicInbox/CodeSearchIdx.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/PublicInbox/CodeSearchIdx.pm b/lib/PublicInbox/CodeSearchIdx.pm
index 4ed5ea64..9ceef16c 100644
--- a/lib/PublicInbox/CodeSearchIdx.pm
+++
Konstantin Ryabitsev wrote:
> On Tue, Nov 14, 2023 at 04:36:29PM +0000, Eric Wong wrote:
> > In any case, kernel.org folks should be able to import missing
> > messages from GNU.org idempotently into lore/qemu-devel without
> > having to resend:
> >
> > https:
Salil Mehta wrote:
> Rest other below link end up in 'Not found'
>
> https://yhbt.net/lore/qemu-devel/20231027150536.3c481...@imammedo.users.ipa.redhat.com/
> https://yhbt.net/lore/qemu-devel/20231027160814.3f47f...@imammedo.users.ipa.redhat.com/
>
Noticed while working on another feature...
---
lib/PublicInbox/CodeSearchIdx.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/PublicInbox/CodeSearchIdx.pm b/lib/PublicInbox/CodeSearchIdx.pm
index 4ed5ea64..9ceef16c 100644
--- a/lib/PublicInbox/CodeSearchIdx.pm
+++
Salil Mehta wrote:
> I have cross confirmed the behavior with other people across companies and
> all of them are having issues in viewing above links. Surprising part is
> these were present at the first instance when the review comments were floated
> by Igor - I can assure you that.
Does my
I don't think this affects anyone, but historically we supported
uppercase characters in newsgroup names since (IIRC) the venerable
INN server was case-sensitive. However, IMAP, POP3 require
lowercase; and our extindex (and WIP -cindex) get confused/broken
with uppercase characters.
So I think
-ggdb3 is already used for g++ and clang, and -pipe is supported
by clang even if it's a no-op. So just use it to speed up g++
since it saves me 30-40ms.
We'll also get rid of the explicit `-O0' since it's the default
for both clang and g++.
---
lib/PublicInbox/XapHelperCxx.pm | 2 +-
1 file
None of our current code relies on it, and I can't imagine it's
something we'd need in the future, actually... This keeps the
door open for relying more on Spawn in TestCommon.
---
lib/PublicInbox/Spawn.pm | 2 +-
t/spawn.t| 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
PublicInbox::IO already gets loaded by PublicInbox::Spawn, so
there's no avoiding it even if we want fast startup time :<
But startup time for this piece will be less relevant in the
near future...
---
lib/PublicInbox/XapHelperCxx.pm | 16 ++--
1 file changed, 6 insertions(+), 10
`stat' can fail due to bugs on our end or ENOMEM, but there's
no autodie support for it. So just die if `unlink' fails, since
the FS wouldn't be usable for tmpfiles in that state, anyways.
---
lib/PublicInbox/Tmpfile.pm | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git
This is shorthand to enabling --associate with the most
aggressive (and time-consuming) options available, starting from
the Unix epoch and having an unlimited window to join on.
---
lib/PublicInbox/CodeSearchIdx.pm | 5 +
script/public-inbox-cindex | 1 +
2 files changed, 6
We'll require an error stream for dump_ibx and dump_roots
commands; they're too important to ignore. Instead of writing
code to provide diagnostics for errors, rely on abort(3) and the
-ggdb3 compiler flag to generate nice core dumps for gdb since
all commands sent to xap_helper are from internal
No need to suffer through an extra dose of slow Perl load times
when we can drive the build in the big parent Perl process and
get the executable path name to pass to spawn directly.
---
lib/PublicInbox/XapClient.pm| 28 ++--
lib/PublicInbox/XapHelperCxx.pm | 9
This will help us notice bugs and system resource limitations
sooner rather than later.
---
lib/PublicInbox/CidxXapHelperAux.pm | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/lib/PublicInbox/CidxXapHelperAux.pm
b/lib/PublicInbox/CidxXapHelperAux.pm
index
I just forgot to use --all with --associate and it wasn't
easily apparent what was wrong. We'll also show some extra
progress while we're at it.
---
lib/PublicInbox/CodeSearchIdx.pm | 26 +-
1 file changed, 21 insertions(+), 5 deletions(-)
diff --git
"window" is probably a better term since it's an inexact thing
to match on.
---
lib/PublicInbox/CodeSearchIdx.pm | 10 +-
script/public-inbox-cindex | 2 +-
2 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/lib/PublicInbox/CodeSearchIdx.pm
Prune can get rid of invalid commits while indexing can add new
candidates for association, so we don't dump coderepo roots for
association until those are squared away. However, we can dump
inbox info since we don't touch inboxes while -cindex is running.
---
lib/PublicInbox/CidxComm.pm |
read_all can be expanded to support FIFOs/pipes/sockets where
read-until-EOF behavior is desired. We can also rely on
wantarray to support splitting on EOL markers, but it's
hard-coded to support only `$/ eq "\n"' since (AFAIK)
it's the only way we use the wantarray form `readline'.
---
We can return an array to allow the caller to omit the internal
`-m' arg entirely. We'll also allow any non-positive values to
mean there's no limit; and we'll defer the "unlimited" case to
the XapHelper implementation. This frees us of having to deal
with mismatches between Perl and Xapian if
We can let these pipes get auto-closed upon leaving the process
subroutine scope.
---
lib/PublicInbox/CodeSearchIdx.pm | 19 +--
1 file changed, 9 insertions(+), 10 deletions(-)
diff --git a/lib/PublicInbox/CodeSearchIdx.pm b/lib/PublicInbox/CodeSearchIdx.pm
index
The C++ version does, so the Perl/XS version should, too;
even if we intentionally avoid using it right now.
---
lib/PublicInbox/XapHelper.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/PublicInbox/XapHelper.pm b/lib/PublicInbox/XapHelper.pm
index 1ee918e3..4157600f
We'll use `kbuf' for the search object key, since we already use
the `fbuf' term in `struct fbuf'. This also adds an extra check
for open_memstream(3) failures in case of ENOMEM.
---
lib/PublicInbox/xap_helper.h | 33 -
1 file changed, 16 insertions(+), 17
We need to have stable filenames and separate compilation
from the linkage stage for ccache to hit. So avoid the use
of a temporary directory and instead rely on a lock file to
guard against parallel builds.
---
lib/PublicInbox/XapHelperCxx.pm | 32 +++-
1 file
refixes=patchid+dfblob
But Perl doesn't ship with getsubopt(3) emulation
out-of-the-box
Eric Wong (18):
cindex: check `say' errors w/ close or ->flush
tmpfile: check `stat' errors, use autodie for unlink
cindex: use `local' for pipes between processes
xap_helper_cxx: use write_file help
We actually need to rely on autodie `close' to check for errors,
since error-checking with `say' is not useful due to perlio
write buffering. We'll also stop relying on `say ... or die'
since it's needless noise.
Fixes: 19f9089343c9 (cindex: drop redundant close on regular FH)
---
I only noticed this while doing a full -cindex --associate with
--associate-date-range=30.years.ago and --associate-max=-1 (no
limit for Xapian) between local mirrors of lore and
git.kernel.org my glibc-based system.
Apparently, glibc requires `optind = 0' to reset getopt(3) in
our workers.
Henrik Grimler wrote:
> Aha, I see, thanks for the explanation! Without the single quotes, and
> after escaping parantheses, lei works as expected.
Good to know.
> For the record, I read some old posts where query was '' quoted, and
> thought it was the way to do it (for example
>
We must use a foreground process to read from terminals
on stdin, otherwise weird things like lost keystrokes and
EIO can happen. So take advantage of ->send_exec_cmd to
spawn `cat' in the same way we spawn MUAs, pagers,
`git config --edit' and `git credential' from script/lei
---
Henrik Grimler wrote:
> Hi Eric,
>
> On Sun, Nov 12, 2023 at 12:10:50AM +0000, Eric Wong wrote:
> > Henrik Grimler wrote:
> > > Hi,
> > >
> > > I recently found out about lei and installed it through archlinux's
> > > package manager
Henrik Grimler wrote:
> Hi,
>
> I recently found out about lei and installed it through archlinux's
> package manager and am trying out queries. When using AND/OR extra
> quotes are inserted in the curl command which messes it up, for
> example:
>
> $ lei q -I https://lore.kernel.org/all/ -o
The whitelist was only used in the early days of its development
and hasn't existed for a while. I've largely forgotten this
thing exists since it's been working well...
---
examples/README.unsubscribe | 9 -
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git
-mda now honors `--help' properly and invocations missing
ORIGINAL_RECIPIENT now fail with EX_NOUSER.
Helped-by: Leah Neukirchen
Link: https://public-inbox.org/meta/87msvlguqu@vuxu.org/
---
script/public-inbox-mda | 7 ++-
t/mda.t | 18 ++
2 files
When learning and injecting new messages ham, we want to avoid
wasting cycles importing the same message into an inbox twice
(once for the To/Cc match and once for the List-Id match). Our
existing %seen hash turned out to be ineffective since
PublicInbox::Inbox refs get re-blessed to
noticed while working on 2.
Eric Wong (4):
learn: fix redundant ham import on dual matches
mda|learn|watch: support dropUniqueUnsubscribe config
mda: fix and test some usage problems
doc: update README.unsubscribe
Documentation/public-inbox-config.pod | 17
Documentation/public-inbox
List-Unsubscribe headers with unique identifiers (such as those
generated by our examples/unsubscribe.milter) should not
end up in public archives. Add a new config knob to strip
List-Unsubscribe headers if they have the
`List-Unsubscribe-Post: List-Unsubscribe=One-Click'
header.
Unfortunately,
Systems with Yama can restrict ptrace(2) (the underlying syscall
used by strace(1)) and make it difficult to test error handling
via error injection. Just skip the tests on such systems since
it's probably not worth the effort to start using prctl(2) to
enable the test on such systems.
---
Konstantin Ryabitsev wrote:
> On Fri, Nov 10, 2023 at 03:09:59AM +0000, Eric Wong wrote:
> > That said, the Atom feeds generated by this RFC includes full
> > messages because that's the easiest way to tie into our existing
> > Atom generation code, so it's currently
Leah Neukirchen wrote:
> Many MTA understand these and map them to sensible SMTP error messages.
>
> Inability to find an inbox results in "5.1.1 user unknown".
> Misformatted messages are rejected with "5.6.0 data format error".
> Unsupported inbox versions are reported as "5.3.5 local
Konstantin Ryabitsev wrote:
> On Thu, Nov 09, 2023 at 02:45:08AM +0000, Eric Wong wrote:
> > This seems like a easy (but WWW-specific) way to get recent
> > topics as suggested by Konstantin. Perhaps an Atom endpoint
> > will also be useful.
>
> Yes, actually th
We've updated all of our users to use Process::IO (and avoiding
tied handles) so the trade-off for using the array context
no longer exists.
---
lib/PublicInbox/Spawn.pm | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/lib/PublicInbox/Spawn.pm b/lib/PublicInbox/Spawn.pm
This matches the behavior we have for multi-message mbox files
since we rely on ->close to detect errors on bad mboxes. This
ensures we'll notice errors reading single messages from stdin.
We'll also start relying more on strace error injection to test
error handling.
---
We can rely on Process::IO->DESTROY to close and reap
in these cases. This is the final step in eliminating
the wantarray invocations of popen_rd (and popen_wr).
---
lib/PublicInbox/LeiInput.pm | 13 +
lib/PublicInbox/LeiRemote.pm | 14 ++
lib/PublicInbox/LeiXSearch.pm
Sometimes seeing that warning is alarming.
---
lib/PublicInbox/LeiMirror.pm | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/lib/PublicInbox/LeiMirror.pm b/lib/PublicInbox/LeiMirror.pm
index 49febe9e..84266d03 100644
--- a/lib/PublicInbox/LeiMirror.pm
+++
We can rely on Process::IO->attached_pid and work towards
simplifying popen_rd.
---
lib/PublicInbox/Xapcmd.pm | 12 +---
1 file changed, 5 insertions(+), 7 deletions(-)
diff --git a/lib/PublicInbox/Xapcmd.pm b/lib/PublicInbox/Xapcmd.pm
index c2b66e69..69f0af43 100644
---
This should allow us to detect shutdown signals in -watch
more quickly and not unnecessarily fail on inconsequential
signals such as SIGWINCH.
---
lib/PublicInbox/NetNNTPSocks.pm | 1 +
lib/PublicInbox/NetReader.pm| 53 +++--
lib/PublicInbox/Watch.pm| 2
In the rare case sendmsg(2) isn't able to send the full amount
(due to buffers >=2GB on Linux), use print + (autodie)close
to send the remainder and retry on EINTR. `substr' should be
able to avoid a large malloc via offsets and CoW on modern Perl.
---
lib/PublicInbox/IPC.pm | 13 +++--
All network connections may fail, so try to emit a helpful
error message instead of attempting to dispatch methods off
`undef'.
---
lib/PublicInbox/LeiLsMailSource.pm | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/lib/PublicInbox/LeiLsMailSource.pm
::IO subclass.
Eric Wong (13):
lei_xsearch: put query in process title for debugging
lei: use cached $daemon_pid when possible
lei: reuse FDs atfork and close explicitly
lei_up: use v5.12
net_nntp_socks: more comments around how it works
lei ls-mail-source: gracefully handle network
This is convoluted as hell but I can't figure out a better way
to make Net::NNTP work with SOCKS.
---
lib/PublicInbox/NetNNTPSocks.pm | 11 ++-
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/lib/PublicInbox/NetNNTPSocks.pm b/lib/PublicInbox/NetNNTPSocks.pm
index
Having queries in the process titles makes it easier to diagnose
stuck queries due to IPC problems. This was used to diagnose
commit e97a30e7624d (lei: fix SIGPIPE on large result sets to pager)).
---
lib/PublicInbox/LeiXSearch.pm | 12 +++-
1 file changed, 7 insertions(+), 5
->lei_daemon_pid can only be called in the top-level daemon
process when $daemon_pid is valid, so avoid a getpid(2) syscall
in those cases.
---
lib/PublicInbox/LEI.pm | 2 +-
lib/PublicInbox/LeiUp.pm | 4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/PublicInbox/LEI.pm
No unicode_strings dependencies here, AFAIK
---
lib/PublicInbox/LeiUp.pm | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/lib/PublicInbox/LeiUp.pm b/lib/PublicInbox/LeiUp.pm
index 0faa180d..9931f017 100644
--- a/lib/PublicInbox/LeiUp.pm
+++ b/lib/PublicInbox/LeiUp.pm
@@
We'll avoid having a redundant STDERR FD open in lei workers,
and some explicit close() on `lei up' sockets reduces the
likelyhood of inadvertantly open FDs causing processes to
linger.
---
lib/PublicInbox/LEI.pm | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git
Konstantin Ryabitsev wrote:
> Hello:
>
> Following the discussion on the ksummit list [1], I wanted to give someone a
> query
> they could use to keep an eye on any new threads. Is there a xapian query that
> can be used to effectively say "return just top-level messages and exclude any
>
When dealing with large search results, we need to deal with
EPIPE not just from the pager, but also EPIPE or ECONNRESET
between lei_xsearch and lei2mail processes.
Without this fix, lei_xsearch processes could linger and get
stuck writing to dead lei2mail processes if a user aborts the
pager
"Michael S. Tsirkin" wrote:
> On Mon, Nov 06, 2023 at 11:49:06AM -0500, Michael S. Tsirkin wrote:
> > On Mon, Nov 06, 2023 at 02:58:18PM +, Salil Mehta wrote:
> > > Hi Michael,
> > > I have noticed something very strange. Review links have been
> > > disappearing from the
> > > Qemu-devel
This was causing `lei q -f text' output to be uncolored on
color-capable terminals.
Fixes: d3c55d072839 (treewide: use ->close to call ProcessIO->CLOSE)
---
lib/PublicInbox/LeiViewText.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/PublicInbox/LeiViewText.pm
Eric Wong wrote:
> --- a/lib/PublicInbox/DS.pm
> +++ b/lib/PublicInbox/DS.pm
> @@ -341,8 +341,8 @@ sub greet {
> my $ev = EPOLLIN;
> my $wbuf;
> if ($sock->can('accept_SSL') && !$sock->accept_SSL) {
> - return CORE::close($sock) i
David Wei wrote:
> Hi,
>
> I have a problem with lei dfn filters. Here is my query:
>
> lei q -o ~/Mail/overlay -I https://lore.kernel.org/all -t '(dfn:net/* OR
> dfn:drivers/net/ethernet/mellanox/mlx5/* OR
> dfn:drivers/net/ethernet/broadcom/bnxt/*) AND tc:net...@vger.kernel.org AND
>
Eric Wong wrote:
> --- a/lib/PublicInbox/WWW.pm
> +++ b/lib/PublicInbox/WWW.pm
> @@ -588,7 +588,7 @@ sub stylesheets_prepare ($$) {
> next;
> };
> my $ctime = 0;
> - my $loc
There's no need to waste optree space on close() statements for
file handles which are (effectively) read-only on their last use
and incapable of error checking in our Perl code (since they're
only read by git).
Let Perl refcounting take care of it so we have less code to
wade through when
Trying to move away from half my code being "or die" statements...
---
lib/PublicInbox/MultiGit.pm | 11 +--
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/lib/PublicInbox/MultiGit.pm b/lib/PublicInbox/MultiGit.pm
index 9074758a..1e8eb47a 100644
---
We have to deal with UTF-8 data for generating patches, so make
it easier to pass Perl utf8 data to git, diff, sdiff, etc. to
avoid "Wide character" warnings.
---
lib/PublicInbox/MailDiff.pm | 3 +--
lib/PublicInbox/SearchIdx.pm | 2 +-
lib/PublicInbox/Spawn.pm | 30
It's easier-to-read and should open the door for us to get rid
of `tie' for ProcessIO without performance penalties for
more frequently-used perlop calls and ability to do `stat' directly
on the object instead of the awkward `tied' thing.
---
lib/PublicInbox/CodeSearchIdx.pm | 6 +++---
We must use eof() combined with close() to detect errors
in situations involving the readline()' op since `readline'
(and most buffered I/O libraries) have weak error detection
support.
This fixes error detection for files opened for read/write
access. The next commit will fix error detection
We can also rely on `say' in Perl 5.10+ to save us the trouble
of printing a newline.
---
lib/PublicInbox/GitCredential.pm | 12 ++--
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/lib/PublicInbox/GitCredential.pm b/lib/PublicInbox/GitCredential.pm
index ae2c..bb225ff3
write_file is a new API which makes setting up config files more
pleasant, while autodie and scalarref redirects (in tests) have
been available for a while, now. So do what we can to reduce
the code burden we have.
---
t/cindex.t| 15 ---
t/extsearch.t | 48
readline () isn't wrapped by autodie, and there's no
way to know if read(2) errors truncated the readline output.
IO::Handle->error isn't reliable on Perl < v5.34.
Thus, combining the `eof' and `close' (combined with autodie) is
the only way we can detect read(2) errors (injected via strace)
when
This will open the door for us to drop `tie' usage from
ProcessIO completely in favor of OO method dispatch. While
OO method dispatches (e.g. `$fh->close') are slower than normal
subroutine calls, it hardly matters in this case since process
teardown is a fairly rare operation and we continue to
This is pretty convenient way to create files for diff
generation in both WWW and lei. The test suite should also be
able to take advantage of it.
---
MANIFEST | 1 +
lib/PublicInbox/IO.pm| 10 +-
lib/PublicInbox/Import.pm| 6 ++
This fixes two major problems with the use of tie for filehandles:
* no way to do fcntl, stat, etc. calls directly on the tied handle,
forcing callers to use the `tied' perlop to access the underlying
IO::Handle
* needing separate classes to handle blocking and non-blocking I/O
As a result,
The IO package seems like a better home for I/O subs than the
Git package. We lose the 60 second read timeout for `git
cat-file --batch-*' processes since it's probably not necessary
given how reliable the code has proven and things would fall
over hard in other ways if the storage device were
FDs are array indices into the kernel, anyways, so we can
take advantage of space savings and speedups because the
majority of FDs a big process has is going to end up in
the array, anyways.
---
lib/PublicInbox/DS.pm | 18 +-
lib/PublicInbox/LeiStoreErr.pm | 2 +-
2
We may start using rand() in the worker someday if we need
to seed a hash function for caching. It saves us some LoC
in the meantime.
---
lib/PublicInbox/XapHelper.pm | 7 +--
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/lib/PublicInbox/XapHelper.pm
switch to sysread eventually to avoid the double-copy overhead
of buffered bulk I/O, anyways. The only place we really benefit
from userspace buffered disk reads is IdxStack, I think...
The new write_file sub in 9/14 seems long overdue.
Eric Wong (14):
xap_helper.pm: use do_fork to Reset and
> Subject: [PATCH] git: reschedule cleanup if activea
There should be no trailing "a" :x
This is necessary to reliably cleanup cat-file processes for
coderepos in long-lived -netd and -httpd processes if they
haven't been accessed in a while.
Followup-to: 33e99002c552 (git: cleanup un-associated coderepo processes)
---
lib/PublicInbox/Git.pm | 3 ++-
1 file changed, 2 insertions(+),
Drop reference counts ASAP in case it saves us some memory
sooner rather than later. This ought to give us more predictable
resource use and ensure OnDestroy callbacks fire sooner.
There's no need to use `local' to clobber the arrayref anymore,
either.
AFAIK, this doesn't fix any known bug, but
We can map all integer FDs to Perl objects once ->ep_wait returns,
so there's no need to play tricks elsewhere to ensure FDs can
be mapped to objects within the same event loop iteration.
---
lib/PublicInbox/DS.pm | 67 ++-
1 file changed, 22 insertions(+),
This allows us support SSL session caching + reuse in the future.
---
lib/PublicInbox/POP3.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/PublicInbox/POP3.pm b/lib/PublicInbox/POP3.pm
index 6d24b17c..06772069 100644
--- a/lib/PublicInbox/POP3.pm
+++
There's no need to waste time nor reach into DS internals to
map FDs to Perl objects, here. LEI.pm has never had to deal
with integer FDs for DirIdle, either.
---
lib/PublicInbox/Watch.pm | 11 ---
1 file changed, 4 insertions(+), 7 deletions(-)
diff --git a/lib/PublicInbox/Watch.pm
Matching existing Perl IO semantics seems like a good idea to
reduce confusion in the future.
We'll also fix some outdated comments and update indentation
to match the rest of our code base since we're far detached from
Danga::Socket at this point.
---
lib/PublicInbox/CidxComm.pm | 2 +-
301 - 400 of 7360 matches
Mail list logo
