* [PATCH] cindex: support sha256 coderepos alongside sha1
@ 2023-04-19 21:54 Eric Wong
2023-04-20 0:53 ` [PATCH 2/1] cindex: limit parallelism of extensions.objectFormat check Eric Wong
0 siblings, 1 reply; 2+ messages in thread
From: Eric Wong @ 2023-04-19 21:54 UTC (permalink / raw)
To: meta
This special support is only needed for --prune at the moment
since the indexing side works on a per-repo basis. There's no
automated tests, yet, but it seems to work well on my sha256
projects when sharing a cindex with sha1 projects.
---
lib/PublicInbox/CodeSearchIdx.pm | 100 +++++++++++++++++++++----------
lib/PublicInbox/Git.pm | 4 +-
lib/PublicInbox/Import.pm | 10 +++-
3 files changed, 79 insertions(+), 35 deletions(-)
diff --git a/lib/PublicInbox/CodeSearchIdx.pm b/lib/PublicInbox/CodeSearchIdx.pm
index 78032c00..54dbf785 100644
--- a/lib/PublicInbox/CodeSearchIdx.pm
+++ b/lib/PublicInbox/CodeSearchIdx.pm
@@ -31,6 +31,7 @@ use PublicInbox::Config qw(glob2re);
use PublicInbox::Spawn qw(spawn popen_rd);
use PublicInbox::OnDestroy;
use PublicInbox::CidxLogP;
+use PublicInbox::Git qw(%OFMT2HEXLEN);
use Socket qw(MSG_EOR);
use Carp ();
our (
@@ -44,7 +45,6 @@ our (
@RDONLY_XDB, # Xapian::Database
@IDX_SHARDS, # clones of self
$MAX_SIZE,
- $TMP_GIT, # PublicInbox::Git object for --prune
$REINDEX, # PublicInbox::SharedKV
@GIT_DIR_GONE, # [ git_dir1, git_dir2 ]
%TO_PRUNE, # (docid => docid) mapping (hash in case of retry_reopen)
@@ -58,6 +58,9 @@ our (
%ACTIVE_GIT_DIR, # GIT_DIR => undef mapping for prune
$IDX_TODO, # [ $git0, $root0, $git1, $root1, ...]
$GIT_TODO, # [ GIT_DIR0, GIT_DIR1, ...]
+ %HEXLEN2TMPGIT, # ((40|64) => PublicInbox::Git for prune)
+ %ALT_FH, # '', or 'sha256' => tmp IO for TMPGIT alternates
+ $TMPDIR, # File::Temp->newdir object
);
# stop walking history if we see >$SEEN_MAX existing commits, this assumes
@@ -701,28 +704,33 @@ sub event_step { # may be requeued via DS
scalar(@cmt) == 1 or warn
"BUG? shard[$self->{shard}] #$PRUNE_CUR has multiple commits: @cmt";
for my $o (@cmt) {
- $TMP_GIT->check_async($o, \&prune_cb,
- [$self, $PRUNE_CUR])
+ $HEXLEN2TMPGIT{length($o)}->check_async($o,
+ \&prune_cb, [$self, $PRUNE_CUR])
}
}
}
- $TMP_GIT->async_wait_all;
+ $_->async_wait_all for (values %HEXLEN2TMPGIT);
cidx_ckpoint($self);
return PublicInbox::DS::requeue($self) if $PRUNE_CUR <= $PRUNE_MAX;
send($PRUNE_OP_P, "prune_done $self->{shard}", MSG_EOR);
$PRUNE_NR //= 0;
progress($self, "prune [$self->{shard}] $PRUNE_NR done");
- $TMP_GIT->cleanup;
- $TMP_GIT = $PRUNE_OP_P = $PRUNE_CUR = $PRUNE_MAX = undef;
- %ACTIVE_GIT_DIR = ();
+ $_->cleanup for (values %HEXLEN2TMPGIT);
+ $PRUNE_OP_P = $PRUNE_CUR = $PRUNE_MAX = undef;
+ undef %ACTIVE_GIT_DIR;
+ undef %HEXLEN2TMPGIT;
}
sub prune_start { # via wq_io_do in IDX_SHARDS
- my ($self, $git_dir, @active_git_dir) = @_;
+ my ($self, $tmpdir, @active_git_dir) = @_;
$PRUNE_CUR = 1;
$PRUNE_OP_P = delete $self->{0} // die 'BUG: no {0} op_p';
%ACTIVE_GIT_DIR = map { $_ => undef } @active_git_dir;
- $TMP_GIT = PublicInbox::Git->new($git_dir); # TMP_GIT copy
+ for my $git_dir (<$tmpdir/*.git>) {
+ my ($hexlen) = ($git_dir =~ m!/hexlen([0-9]+)\.git\z!);
+ $hexlen or die "BUG: no hexlen in $git_dir";
+ $HEXLEN2TMPGIT{$hexlen} = PublicInbox::Git->new($git_dir);
+ }
$self->begin_txn_lazy;
$PRUNE_MAX = $self->{xdb}->get_lastdocid // 1;
event_step($self);
@@ -750,24 +758,6 @@ sub parent_quit {
warn "# SIG$_[0] received, quitting...\n";
}
-sub init_tmp_git_dir ($) {
- my ($self) = @_;
- require File::Temp;
- require PublicInbox::Import;
- my $tmp = File::Temp->newdir('cidx-all-git-XXXX', TMPDIR => 1);
- PublicInbox::Import::init_bare("$tmp", 'cidx-all');
- my $f = "$tmp/objects/info/alternates";
- open my $fh, '>', $f or die "open($f): $!";
- my $o;
- for (@{$self->{git_dirs}}) { # TODO: sha256 check?
- $o = $_.'/objects';
- say $fh $o if -d $o;
- }
- close $fh or die "close($f): $!";
- $TMP_GIT = PublicInbox::Git->new("$tmp");
- $TMP_GIT->{-tmp} = $tmp;
-}
-
sub prep_umask ($) {
my ($self) = @_;
if ($self->{-cidx_internal}) { # respect core.sharedRepository
@@ -789,16 +779,60 @@ sub prep_umask ($) {
}
}
-sub start_prune ($) {
+sub prep_alternate { # awaitpid callback for config extensions.objectFormat
+ my ($pid, $objdir, $out, $send_prune) = @_;
+ my $status = $? >> 8;
+ my $fmt;
+ if ($status == 1) { # unset, default is '' (SHA-1)
+ $fmt = 'sha1';
+ } elsif ($status == 0) {
+ seek($out, 0, SEEK_SET) or die "seek: $!";
+ chomp($fmt = <$out> // 'sha1');
+ } else {
+ return warn("git config \$?=$? for objdir=$objdir");
+ }
+ my $hexlen = $OFMT2HEXLEN{$fmt} // return warn <<EOM;
+E: ignoring objdir=$objdir, unknown extensions.objectFormat=$fmt
+EOM
+ unless ($ALT_FH{$fmt}) {
+ my $git_dir = "$TMPDIR/hexlen$hexlen.git";
+ PublicInbox::Import::init_bare($git_dir, 'cidx-all', $fmt);
+ my $f = "$git_dir/objects/info/alternates";
+ open $ALT_FH{$fmt}, '>', $f or die "open($f): $!";
+ }
+ say { $ALT_FH{$fmt} } $out or die "say: $!";
+ # send_prune fires on the last one
+}
+
+sub init_prune ($) {
my ($self) = @_;
return (@$PRUNE_DONE = map { 1 } @IDX_SHARDS) if !$self->{-opt}->{prune};
- init_tmp_git_dir($self);
+
+ require File::Temp;
+ require PublicInbox::Import;
+ $TMPDIR = File::Temp->newdir('cidx-all-git-XXXX', TMPDIR => 1);
+ my $send_prune = PublicInbox::OnDestroy->new($$, \&send_prune, $self);
+ my $cmd = [ 'git', undef, 'config', 'extensions.objectFormat' ];
+ for (@{$self->{git_dirs}}) {
+ my $o = $_.'/objects';
+ next if !-d $o;
+ $cmd->[1] = "--git-dir=$_";
+ open my $out, '+>', undef or die "open(tmp): $!";
+ my $pid = spawn($cmd, undef, { 1 => $out });
+ awaitpid($pid, \&prep_alternate, $o, $out, $send_prune);
+ }
+}
+
+sub send_prune { # OnDestroy when `git config extensions.objectFormat' are done
+ my ($self) = @_;
+ for (values %ALT_FH) { close $_ or die "close: $!" }
+ %ALT_FH = ();
my @active_git_dir = (@{$self->{git_dirs}}, @GIT_DIR_GONE);
my ($c, $p) = PublicInbox::PktOp->pair;
$c->{ops}->{prune_done} = [ $self ];
for my $s (@IDX_SHARDS) {
$s->wq_io_do('prune_start', [ $p->{op_p} ],
- $TMP_GIT->{git_dir}, @active_git_dir)
+ "$TMPDIR", @active_git_dir)
}
}
@@ -812,8 +846,8 @@ sub cidx_run { # main entry point
local $LIVE = {};
local $PRUNE_DONE = [];
local $IDX_TODO = [];
- local ($DO_QUIT, $TMP_GIT, $REINDEX, $TXN_BYTES, @GIT_DIR_GONE,
- $GIT_TODO, $REPO_CTX);
+ local ($DO_QUIT, $REINDEX, $TXN_BYTES, @GIT_DIR_GONE,
+ $GIT_TODO, $REPO_CTX, %ALT_FH, $TMPDIR, %HEXLEN2TMPGIT);
local $BATCH_BYTES = $self->{-opt}->{batch_size} //
$PublicInbox::SearchIdx::BATCH_BYTES;
local @IDX_SHARDS = cidx_init($self);
@@ -859,7 +893,7 @@ sub cidx_run { # main entry point
local $LIVE_JOBS = $self->{-opt}->{jobs} ||
PublicInbox::IPC::detect_nproc() || 2;
local @RDONLY_XDB = $self->xdb_shards_flat;
- start_prune($self);
+ init_prune($self);
scan_git_dirs($self) if $self->{-opt}->{scan} // 1;
local @PublicInbox::DS::post_loop_do = (\&shards_active);
diff --git a/lib/PublicInbox/Git.pm b/lib/PublicInbox/Git.pm
index 3108ed85..61ba8aa1 100644
--- a/lib/PublicInbox/Git.pm
+++ b/lib/PublicInbox/Git.pm
@@ -22,7 +22,9 @@ use IO::Poll qw(POLLIN);
use Carp qw(croak carp);
use PublicInbox::SHA ();
use PublicInbox::DS qw(awaitpid);
-our @EXPORT_OK = qw(git_unquote git_quote);
+our %HEXLEN2SHA = (40 => 1, 64 => 256);
+our %OFMT2HEXLEN = (sha1 => 40, sha256 => 64);
+our @EXPORT_OK = qw(git_unquote git_quote %HEXLEN2SHA %OFMT2HEXLEN);
our $PIPE_BUFSIZ = 65536; # Linux default
our $in_cleanup;
our $RDTIMEO = 60_000; # milliseconds
diff --git a/lib/PublicInbox/Import.pm b/lib/PublicInbox/Import.pm
index 39719bcb..59462e9a 100644
--- a/lib/PublicInbox/Import.pm
+++ b/lib/PublicInbox/Import.pm
@@ -461,13 +461,21 @@ my @INIT_FILES = ('HEAD' => undef, # filled in at runtime
EOC
sub init_bare {
- my ($dir, $head) = @_; # or self
+ my ($dir, $head, $fmt) = @_; # or self
$dir = $dir->{git}->{git_dir} if ref($dir);
require File::Path;
File::Path::make_path(map { $dir.$_ } qw(/objects/info /refs/heads));
$INIT_FILES[1] //= 'ref: '.default_branch."\n";
my @fn_contents = @INIT_FILES;
$fn_contents[1] = "ref: refs/heads/$head\n" if defined $head;
+ $fn_contents[3] = <<EOM if defined($fmt) && $fmt ne 'sha1';
+[core]
+ repositoryFormatVersion = 1
+ filemode = true
+ bare = true
+[extensions]
+ objectFormat = $fmt
+EOM
while (my ($fn, $contents) = splice(@fn_contents, 0, 2)) {
my $f = $dir.'/'.$fn;
next if -f $f;
^ permalink raw reply related [flat|nested] 2+ messages in thread
* [PATCH 2/1] cindex: limit parallelism of extensions.objectFormat check
2023-04-19 21:54 [PATCH] cindex: support sha256 coderepos alongside sha1 Eric Wong
@ 2023-04-20 0:53 ` Eric Wong
0 siblings, 0 replies; 2+ messages in thread
From: Eric Wong @ 2023-04-20 0:53 UTC (permalink / raw)
To: meta
We can't safely spawn all `git config' processes of every
indexed git directory at once due to system resource limits
(RLIMIT_NPROC, RLIMIT_NOFILE). So queue them up and limit
parallelism that way.
---
lib/PublicInbox/CodeSearchIdx.pm | 33 +++++++++++++++++++++-----------
1 file changed, 22 insertions(+), 11 deletions(-)
diff --git a/lib/PublicInbox/CodeSearchIdx.pm b/lib/PublicInbox/CodeSearchIdx.pm
index 54dbf785..97123133 100644
--- a/lib/PublicInbox/CodeSearchIdx.pm
+++ b/lib/PublicInbox/CodeSearchIdx.pm
@@ -61,6 +61,7 @@ our (
%HEXLEN2TMPGIT, # ((40|64) => PublicInbox::Git for prune)
%ALT_FH, # '', or 'sha256' => tmp IO for TMPGIT alternates
$TMPDIR, # File::Temp->newdir object
+ @PRUNE_QUEUE, # GIT_DIRs to prepare for pruning
);
# stop walking history if we see >$SEEN_MAX existing commits, this assumes
@@ -779,9 +780,11 @@ sub prep_umask ($) {
}
}
-sub prep_alternate { # awaitpid callback for config extensions.objectFormat
+sub prep_alternate_end { # awaitpid callback for config extensions.objectFormat
my ($pid, $objdir, $out, $send_prune) = @_;
my $status = $? >> 8;
+ my $next_dir = shift(@PRUNE_QUEUE);
+ prep_alternate_start($next_dir, $send_prune) if defined($next_dir);
my $fmt;
if ($status == 1) { # unset, default is '' (SHA-1)
$fmt = 'sha1';
@@ -801,7 +804,20 @@ EOM
open $ALT_FH{$fmt}, '>', $f or die "open($f): $!";
}
say { $ALT_FH{$fmt} } $out or die "say: $!";
- # send_prune fires on the last one
+}
+
+sub prep_alternate_start {
+ my ($git_dir, $send_prune) = @_;
+ my $o = $git_dir.'/objects';
+ while (!-d $o) {
+ $git_dir = shift(@PRUNE_QUEUE) // return
+ $o = $git_dir.'/objects';
+ }
+ my $cmd = [ 'git', "--git-dir=$git_dir",
+ qw(config extensions.objectFormat) ];
+ open my $out, '+>', undef or die "open(tmp): $!";
+ my $pid = spawn($cmd, undef, { 1 => $out });
+ awaitpid($pid, \&prep_alternate_end, $o, $out, $send_prune);
}
sub init_prune ($) {
@@ -812,14 +828,9 @@ sub init_prune ($) {
require PublicInbox::Import;
$TMPDIR = File::Temp->newdir('cidx-all-git-XXXX', TMPDIR => 1);
my $send_prune = PublicInbox::OnDestroy->new($$, \&send_prune, $self);
- my $cmd = [ 'git', undef, 'config', 'extensions.objectFormat' ];
- for (@{$self->{git_dirs}}) {
- my $o = $_.'/objects';
- next if !-d $o;
- $cmd->[1] = "--git-dir=$_";
- open my $out, '+>', undef or die "open(tmp): $!";
- my $pid = spawn($cmd, undef, { 1 => $out });
- awaitpid($pid, \&prep_alternate, $o, $out, $send_prune);
+ @PRUNE_QUEUE = @{$self->{git_dirs}};
+ for (1..$LIVE_JOBS) {
+ prep_alternate_start(shift(@PRUNE_QUEUE) // last, $send_prune);
}
}
@@ -846,7 +857,7 @@ sub cidx_run { # main entry point
local $LIVE = {};
local $PRUNE_DONE = [];
local $IDX_TODO = [];
- local ($DO_QUIT, $REINDEX, $TXN_BYTES, @GIT_DIR_GONE,
+ local ($DO_QUIT, $REINDEX, $TXN_BYTES, @GIT_DIR_GONE, @PRUNE_QUEUE,
$GIT_TODO, $REPO_CTX, %ALT_FH, $TMPDIR, %HEXLEN2TMPGIT);
local $BATCH_BYTES = $self->{-opt}->{batch_size} //
$PublicInbox::SearchIdx::BATCH_BYTES;
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2023-04-20 0:53 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-04-19 21:54 [PATCH] cindex: support sha256 coderepos alongside sha1 Eric Wong
2023-04-20 0:53 ` [PATCH 2/1] cindex: limit parallelism of extensions.objectFormat check Eric Wong
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).