Split out tex-fontinst

(markd)

Add tex-hyphen-english 2014

Additional hyphenation patterns for American and British
English in ASCII encoding.  The American English patterns
(usenglishmax) greatly extend the standard patterns from Knuth
to find many additional hyphenation points.  British English
hyphenation is completely different from US English, so has its
own set of patterns.

(markd)

Note conflict with teTeX3-texmf.

(markd)

Update to tex-lambda 2014

changes - removal of two files.

(markd)

Refine the TeX build dependencies.

(markd)

Refine the dependencies. Bump PKGREVISION.

(markd)

+devel/p5-Find-Lib version 1.04
+devel/p5-MetaCPAN-Client version 1.013000
+devel/p5-Search-Elasticsearch version 1.19
+net/p5-Test-RequiresInternet version 0.04
+www/p5-Any-URI-Escape version 0.01
+www/p5-Hijk version 0.20
+www/p5-WWW-Mechanize-Cached version 1.48

(mef)

Updated textproc/p5-Text-Format to 0.59
Updated textproc/p5-Text-Hunspell to 2.10

(mef)

Update to 2.10
--------------
2.10    2015/03/26 11:24:00 Europe/Oslo (cosimo)

        Fixed GH#5 (http://github.com/cosimo/perl5-text-hunspell/issues/5).
        Should make Text::Hunspell installable without pain from CPAN and
        cpanminus clients. The problem was in the undeclared use of
        ExtUtils::PkgConfig, which needs to be known even before running
        the configure (Makefile.PL) step. Thanks to Pavel Shaydo for reporting
        the problem.

(mef)

Update to 0.59
--------------
0.59    2014-02-01
    - Add the missing README to the distribution (CPANTS).
    - Add a LICENSE file (CPANTS).
    - Add a "=head1 LICENSE" section to the POD (CPANTS).
    - Minimum version of perl in META.yml (CPANTS).
    - Add "use warnings;" (CPANTS).
    - Add t/cpan-changes.t and t/style-trailing-space.t .

(mef)

Updated devel/p5-Git-CPAN-Patch to 2.0.4
Updated devel/p5-Test-SharedFork to 0.29

(mef)

Update 0.29 to 0.31
-------------------
0.31 2015-04-28T20:25:31Z
    - Fix test that accidentally got commited last release

0.30 2015-04-28T20:19:58Z
    - Minor changes to support newer Test::Stream versions

(mef)

(pkgsrc)
- Convert PERL5_MODULE_TYPE to default
- Add 5 of BUILD_DEPENDS for make test

(upstream)
- update to 2.0.4
---------------
2.0.4 2015-04-19
[BUG FIXES]
- Tests were failing because of space-sensitivity. (lharey, GH#21)

[ENHANCEMENTS]
- Move to MetaCPAN::Client. (GH#19)

[STATISTICS]
- code churn: 14 files changed, 233 insertions(+), 210 deletions(-)

(mef)

Added devel/p5-Find-Lib version 1.04
Added www/p5-Hijk version 0.20
Added devel/p5-Search-Elasticsearch version 1.19
Added net/p5-Test-RequiresInternet version 0.04
Added www/p5-Any-URI-Escape version 0.01
Added www/p5-WWW-Mechanize-Cached version 1.48
Added devel/p5-MetaCPAN-Client version 1.013000

(mef)

Clear pkglint flags, newline at the bottom of DESCR, no A in comment line.

(mef)

Import p5-MetaCPAN-Client-1.013000 as devel/p5-MetaCPAN-Client.

This is a hopefully-complete API-compliant client to MetaCPAN
(https://metacpan.org) with DWIM capabilities, to make your life
easier.CVS: ----------------------------------------------------------------------

(mef)

Import p5-WWW-Mechanize-Cached-1.48 as www/p5-WWW-Mechanize-Cached.

Uses the Cache::Cache hierarchy by default to implement a caching
Mech. This lets one perform repeated requests without hammering a
server impolitely.
  Please note that Cache::Cache has been superceded by CHI, but the
default has not been changed here for reasons of backwards
compatibility. For this reason, you are encouraged to provide your own
CHI caching object to override the default.

(mef)

Import p5-Any-URI-Escape-0.01 as www/p5-Any-URI-Escape.

URI::Escape is great, but URI::Escape::XS is faster. This module loads
URI::Escape::XS and imports the two most common methods if XS is
installed.

(mef)

Import p5-Test-RequiresInternet-0.04 as net/p5-Test-RequiresInternet.

This module is intended to easily test network connectivity before
functional tests begin to non-local Internet resources. It does not
require any modules beyond those supplied in core Perl.

(mef)

Import p5-Search-Elasticsearch-1.19 as devel/p5-Search-Elasticsearch.

Search::Elasticsearch is the official Perl client for Elasticsearch,
supported by elasticsearch.com. Elasticsearch itself is a flexible and
powerful open source, distributed real-time search and analytics
engine for the cloud. You can read more about it on elasticsearch.org.

(mef)

Backout bumping PKGREVISON of ruby200-base.

(taca)

Backout previous commit since ruby200-base dose not need this patch.

(taca)

Note update of ruby*-base packages:

lang/ruby193-base 1.9.3p551nb2
lang/ruby200-base 2.0.0p645nb1
lang/ruby21-base 2.1.6nb1
lang/ruby22-base 2.2.2nb1

(taca)

Avoid to generate empty command line on some platforms:

$(DLLIB): $(OBJS) Makefile
...
$(Q) $(POSTLINK)

And POSTLINK is empty macro.  In such case, GNU make ignore empty command
line but BSD make tries to execute it and causes error.

Bump PKGREVISION.

(taca)

Avoid to generate empty command line on some platforms:

$(DLLIB): $(OBJS) Makefile
...
$(Q) $(POSTLINK)

And POSTLINK is empty macro.  In such case, GNU make ignore empty command
line but BSD make tries to execute it and causes error.

Bump PKGREVISION.

(taca)

Avoid to generate empty command line on some platforms:

$(DLLIB): $(OBJS) Makefile
...
$(Q) $(POSTLINK)

And POSTLINK is empty macro.  In such case, GNU make ignore empty command
line but BSD make tries to execute it and causes error.

Bump PKGREVISION.

(taca)

Avoid to generate empty command line on some platforms:

$(DLLIB): $(OBJS) Makefile
...
$(Q) $(POSTLINK)

And POSTLINK is empty macro.  In such case, GNU make ignore empty command
line but BSD make tries to execute it and causes error.

Bump PKGREVISION.

(taca)

Import p5-Hijk-0.20 as www/p5-Hijk.

Hijk is a fast & minimal low-level HTTP client intended to be used
where you control both the client and the server, e.g. for talking to
some internal service from a frontend user-facing web application.

It is NOT a general HTTP user agent, it doesn't support redirects,
proxies, SSL and any number of other advanced HTTP features like (in
roughly descending order of feature completeness) LWP::UserAgent,
WWW::Curl, HTTP::Tiny, HTTP::Lite or Furl. This library is basically
one step above manually talking HTTP over sockets.

Having said that it's lightning fast and extensively used in
production at Booking.com where it's used as the go-to transport layer
for talking to internal services. It uses non-blocking sockets and
correctly handles all combinations of connect/read timeouts and other
issues you might encounter from various combinations of parts of your
system going down or becoming otherwise unavailable.

(mef)

Import p5-Find-Lib-1.04 as devel/p5-Find-Lib.

The purpose of this module is to replace
    use FindBin;
    use lib "$FindBin::Bin/../bootstrap/lib";
with something shorter. This is specially useful if your project has a
lot of scripts (For instance tests scripts).

(mef)

pullup #4690

(hiramatsu)

Pullup ticket #4690 - requested by joerg
sysutils/dvdisaster: build fix

Revisions pulled up:
- sysutils/dvdisaster/Makefile                                  1.32
- sysutils/dvdisaster/distinfo                                  1.8
- sysutils/dvdisaster/patches/patch-md5.c                      1.1
- sysutils/dvdisaster/patches/patch-scripts_bash-based-configure 1.1
- sysutils/dvdisaster/patches/patch-tools_codec.c              1.1
- sysutils/dvdisaster/patches/patch-tools_decimate.c            1.1
- sysutils/dvdisaster/patches/patch-tools_memory.c              1.1
- sysutils/dvdisaster/patches/patch-tools_pngio.c              1.1
- sysutils/dvdisaster/patches/patch-tools_pngpack.c            1.1

---
  Module Name: pkgsrc
  Committed By:        joerg
  Date:                Sun Apr 26 13:30:59 UTC 2015

  Modified Files:
        pkgsrc/sysutils/dvdisaster: Makefile distinfo
  Added Files:
        pkgsrc/sysutils/dvdisaster/patches: patch-md5.c
            patch-scripts_bash-based-configure patch-tools_codec.c
            patch-tools_decimate.c patch-tools_memory.c patch-tools_pngio.c
            patch-tools_pngpack.c

  Log Message:
  Drop GCC check. Fix missing includes. Fix memset call. Bump revision.

(hiramatsu)

One more BUILD_DEPENDS (p2-File-Whick) for make test

(mef)

Updated devel/gitolite to 3.6.3

(ryoon)

Update to 3.6.3

* Use GITHUB framework.

Changelog:
2015-04-26  v3.6.3  allow limited use of 'git config' using the new 'config'
                    command

                    accept openssh 6.8's new fingerprint output format

                    (finally!) allow limited symlinks within ~/repositories;
                    see commit 8e36230 for details

                    perms command now lists available roles

                    minor backward compat breakage: 'perms -l repo' no longer
                    works; see 'perms -h' for new usage

                    allow gitolite-shell to be used as $SHELL (experts only;
                    no support, no docs; see commit 9cd1e37 for details)

                    help with 'git push --signed' using a post-receive hook to
                    adopt push certs into 'refs/push-certs'; for details see
                    contrib/hooks/repo-specific/save-push-signatures

                    new 'transparent proxy' feature for git repos; see
                    src/lib/Gitolite/Triggers/TProxy.pm for details

(ryoon)

Updated textproc/source-highlight to 3.1.8

(mef)

Update to 3.1.8  (from the file NEWS)
---------------
Version 3.1.8

  * fixed some compilation bugs in some systems
  * fixed tcl.lang
  * fixed bash.lang
  * New language definition files:
  * added coffeescript
  * added Cucumber / feature
  * added Go
  * added Groovy
  * added Json
  * added Zsh
  * New output formats:
  * added HTML5
  * added S-expressions
  * added Groff

(mef)

Pullup tickets #4697, #4698, #4699, #4703 and #4705.

(tron)

Pullup ticket #4705 - requested by spz
www/curl: security update

Revisions pulled up:
- www/curl/Makefile                                            1.147
- www/curl/PLIST                                                1.48
- www/curl/distinfo                                            1.103
- www/curl/patches/patch-aa                                    1.30

---
  Module Name: pkgsrc
  Committed By: jperkin
  Date: Wed Apr 22 14:35:21 UTC 2015

  Modified Files:
  pkgsrc/www/curl: Makefile PLIST distinfo
  pkgsrc/www/curl/patches: patch-aa

  Log Message:
  Update to curl-7.42.0.

  This release includes the following changes:

    o openssl: show the cipher selection to use in verbose text
    o gtls: implement CURLOPT_CERTINFO
    o add CURLOPT_SSL_FALSESTART option (darwinssl and NSS)
    o curl: add --false-start option
    o add CURLOPT_PATH_AS_IS
    o curl: add --path-as-is option
    o curl: create output file on successful download of an empty file

  This release includes the following bugfixes:

    o ConnectionExists: for NTLM re-use, require credentials to match
    o cookie: cookie parser out of boundary memory access
    o fix_hostname: zero length host name caused -1 index offset
    o http_done: close Negotiate connections when done
    o sws: timeout idle CONNECT connections
    o nss: improve error handling in Curl_nss_random()
    o nss: do not skip Curl_nss_seed() if data is NULL
    o curl-config.in: eliminate double quotes around CURL_CA_BUNDLE
    o http2: move lots of verbose output to be debug-only
    o dist: add extern-scan.pl to the tarball
    o http2: return recv error on unexpected EOF
    o build: Use default RandomizedBaseAddress directive in VC9+ project files
    o build: Removed DataExecutionPrevention directive from VC9+ project files
    o tool: Updated the warnf() function to use the GlobalConfig structure
    o http2: Return error if stream was closed with other than NO_ERROR
    o mprintf.h: remove #ifdef CURLDEBUG
    o libtest: fixed linker errors on msvc
    o tool: use ENABLE_CURLX_PRINTF instead of _MPRINTF_REPLACE
    o curl.1: fix "The the" typo
    o cmake: handle build definitions CURLDEBUG/DEBUGBUILD
    o openssl: remove all uses of USE_SSLEAY
    o multi: fix memory-leak on timeout (regression)
    o curl_easy_setopt.3: added CURLOPT_SSL_VERIFYSTATUS
    o metalink: add some error checks
    o TLS: make it possible to enable ALPN/NPN without HTTP/2
    o http2: use CURL_HTTP_VERSION_* symbols instead of NPN_*
    o conncontrol: only log changes to the connection bit
    o multi: fix *getsock() with CONNECT
    o symbols.pl: handle '-' in the deprecated field
    o MacOSX-Framework: use @rpath instead of @executable_path
    o GnuTLS: add support for CURLOPT_CAPATH
    o GnuTLS: print negotiated TLS version and full cipher suite name
    o GnuTLS: don't print double newline after certificate dates
    o memanalyze.pl: handle free(NULL)
    o proxy: re-use proxy connections (regression)
    o mk-ca-bundle: Don't report SHA1 numbers with "-q"
    o http: always send Host: header as first header
    o openssl: sort ciphers to use based on strength
    o openssl: use colons properly in the ciphers list
    o http2: detect premature close without data transfered
    o hostip: Fix signal race in Curl_resolv_timeout
    o closesocket: call multi socket cb on close even with custom close
    o mksymbolsmanpage.pl: use std header and generate better nroff header
    o connect: Fix happy eyeballs logic for IPv4-only builds
    o curl_easy_perform.3: remove superfluous close brace from example
    o HTTP: don't use Expect: headers when on HTTP/2
    o Curl_sh_entry: remove unused 'timestamp'
    o docs/libcurl: makefile portability fix
    o mkhelp: Remove trailing carriage return from every line of input
    o nss: explicitly tell NSS to disable NPN/ALPN when libcurl disables it
    o curl_easy_setopt.3: added a few missing options
    o metalink: fix resource leak in OOM
    o axtls: version 1.5.2 now requires that config.h be manually included
    o HTTP: don't switch to HTTP/2 from 1.1 until we get the 101
    o cyassl: detect the library as renamed wolfssl
    o CURLOPT_HTTPHEADER.3: add a "SECURITY CONCERNS" section
    o CURLOPT_URL.3: Added "SECURITY CONCERNS
    o openssl: try to avoid accessing OCSP structs when possible
    o test938: added missing closing tags
    o testcurl: Allow '=' in values given on command line
    o tests/certs: added make target to rebuild certificates
    o tests/certs: rebuild certificates with modified key usage bits
    o gtls: avoid uninitialized variable
    o gtls: dereferencing NULL pointer
    o gtls: add check of return code
    o test1513: eliminated race condition in test run
    o dict: rename byte to avoid compiler shadowed declaration warning
    o curl_easy_recv/send: make them work with the multi interface
    o vtls: fix compile with --disable-crypto-auth but with SSL
    o openssl: adapt to ASN1/X509 things gone opaque in 1.1
    o openssl: verifystatus: only use the OCSP work-around <= 1.0.2a
    o curl_memory: make curl_memory.h the second-last header file loaded
    o testcurl.pl: add the --notes option to supply more info about a build
    o cyassl: If wolfSSL then identify as such in version string
    o cyassl: Check for invalid length parameter in Curl_cyassl_random
    o cyassl: default to highest possible TLS version
    o Curl_ssl_md5sum: return CURLcode (fixes OOM)
    o polarssl: remove dead code
    o polarssl: called mbedTLS in 1.3.10 and later
    o globbing: fix step parsing for character globbing ranges
    o globbing: fix url number calculation when using range with step
    o multi: on a request completion, check all CONNECT_PEND transfers
    o build: link curl to openssl libraries when openssl support is enabled
    o url: Don't accept CURLOPT_SSLVERSION unless USE_SSL is defined
    o vtls: Don't accept unknown CURLOPT_SSLVERSION values
    o build: Fix libcurl.sln erroneous mixed configurations
    o cyassl: remove undefined reference to CyaSSL_no_filesystem_verify
    o cyassl: add SSL context callback support for CyaSSL
    o tool: only set SSL options if SSL is enabled
    o multi: remove_handle: move pending connections
    o configure: Use KRB5CONFIG for krb5-config
    o axtls: add timeout within Curl_axtls_connect
    o CURLOPT_HTTP200ALIASES.3: Mainly SHOUTcast servers use "ICY 200"
    o cyassl: Fix library initialization return value
    o cookie: handle spaces after the name in Set-Cookie
    o http2: Fix missing nghttp2_session_send call in Curl_http2_switched
    o cyassl: Fix certificate load check
    o build-openssl.bat: Fix mixed line endings
    o checksrc.bat: Check lib\vtls source
    o DNS: fix refreshing of obsolete dns cache entries
    o CURLOPT_RESOLVE: actually implement removals
    o checksrc.bat: quotes to support an SRC_DIR with spaces
    o cyassl: Remove 'Connecting to' message from cyassl_connect_step2
    o cyassl: Use CYASSL_MAX_ERROR_SZ for error buffer size
    o lib/transfer.c: Remove factor of 8 from sleep time calculation
    o lib/makefile.m32: add missing libs to build libcurl.dll
    o build: Generate source prerequisites for Visual Studio in generate.bat
    o cyassl: Include the CyaSSL build config
    o firefox-db2pem: fix wildcard to find Firefox default profile
    o BUGS: refer to the github issue tracker now as primary
    o vtls_openssl: improve several certificate error messages
    o cyassl: Add support for TLS extension SNI
    o parsecfg: do not continue past a zero termination
    o configure --with-nss=PATH: query pkg-config if available
    o configure --with-nss: drop redundant if statement
    o cyassl: Fix include order
    o HTTP: fix PUT regression with Negotiate
    o curl_version_info.3: fixed the 'protocols' variable type

(tron)

Pullup ticket #4703 - requested by spz
databases/gnats: security update

Revisions pulled up:
- databases/gnats/Makefile                                      1.38-1.39
- databases/gnats/PLIST                                        1.13
- databases/gnats/distinfo                                      1.15
- databases/gnats/patches/patch-aa                              1.5
- databases/gnats/patches/patch-ab                              deleted
- databases/gnats/patches/patch-ac                              deleted
- databases/gnats/patches/patch-ad                              deleted
- databases/gnats/patches/patch-ae                              deleted
- databases/gnats/patches/patch-ah                              deleted
- databases/gnats/patches/patch-configure                      1.1
- databases/gnats/patches/patch-gnats_configure                deleted
- databases/gnats/patches/patch-gnats_ds-file_mkcat.sh          1.1
- databases/gnats/patches/patch-gnats_ds-file_mkdb.sh          1.1
- databases/gnats/patches/patch-gnats_ds-file_rmcat.sh          1.1
- databases/gnats/patches/patch-gnats_edit-pr.sh                1.1
- databases/gnats/patches/patch-gnats_gnats_h                  1.2
- databases/gnats/patches/patch-makefile_in                    deleted

---
  Module Name: pkgsrc
  Committed By: spz
  Date: Sun Apr 19 19:35:00 UTC 2015

  Modified Files:
  pkgsrc/databases/gnats: Makefile PLIST distinfo
  pkgsrc/databases/gnats/patches: patch-aa patch-gnats_gnats_h
  Added Files:
  pkgsrc/databases/gnats/patches: patch-configure
      patch-gnats_ds-file_mkcat.sh patch-gnats_ds-file_mkdb.sh
      patch-gnats_ds-file_rmcat.sh patch-gnats_edit-pr.sh
  Removed Files:
  pkgsrc/databases/gnats/patches: patch-ab patch-ac patch-ad patch-ae
      patch-ah patch-gnats_configure patch-makefile_in

  Log Message:
  update to gnats 4.2.0

  upstream Changes highlights:
    - GNU v3 license update
    - Buid now uses automake, and code has been adjusted to compile
      properly in various environments (tested with gcc 2.95, 3.4, 4.8,
      4.9, clang 3.1, 3.4, and several different versions of Linux,
      FreeBSD, Solaris, OS X)
    - New --global-databases-file command-line option for gnatsd and other programs
    - New host/port/user/passwd command-line options for send-pr
    - Many bugfixes, including:
    - FreeBSD-SA-05:20.cvsbug security advisory (affected send-pr)
    - Fixed PR#484: "send-pr --request-id" sends just a template
    - Fixed PR#501: Safer open/fopen for gen-index (no longer installing as suid)
    - Fixed PR#504: gnatsd crashes when trying to lock a nonexistent PR
    - Fixed PR#508: Emacs interface gnats-change-database fails when
      gnats-port isn't set
    - Fixed PR#691: Portability problem with sed and tr
    - Fixed PR#746: install-sid is a bash script (and not /bin/sh)
    - Fixed PR#767: edit-pr returns incorect exit status
    - Fixed PR#768: make install DESTDIR=/anywhere fails
    - Fixed PR#771: "EXPR" queries with spaces sometimes fail
    - Fixed PR#773: using responsible alias as email address for edit-pr
    - "invalid query format" bug for queries that included whitespace
    - edit-pr will now correctly exit with an error if it cannot lock the PR
    - edit-pr will now use the email alias in the responsibles field
    - Solaris 7/8/10 compilation fixed

---
  Module Name: pkgsrc
  Committed By: spz
  Date: Fri Apr 24 06:45:28 UTC 2015

  Modified Files:
  pkgsrc/databases/gnats: Makefile

  Log Message:
  pick up changes in gnats' configure arguments

(tron)

Mark as not "MAKE_JOBS_SAFE".

(tron)

Pullup ticket #4699 - requested by spz
sysutils/xenkernel45: security patch

Revisions pulled up:
- sysutils/xenkernel45/Makefile                                1.7
- sysutils/xenkernel45/distinfo                                1.6
- sysutils/xenkernel45/patches/patch-CVE-2015-2751              1.1

---
  Module Name: pkgsrc
  Committed By: spz
  Date: Sun Apr 19 15:02:12 UTC 2015

  Modified Files:
  pkgsrc/sysutils/xenkernel45: Makefile distinfo
  Added Files:
  pkgsrc/sysutils/xenkernel45/patches: patch-CVE-2015-2751

  Log Message:
  adding upstream's patch for
  XSA-127 Certain domctl operations may be abused to lock up the host

(tron)

Pullup ticket #4698 - requested by spz
Pullup ticket #4698 - requested by spz
sysutils/xenkernel41: security patch
sysutils/xenkernel42: security patch
sysutils/xenkernel45: security patch

Revisions pulled up:
- sysutils/xenkernel41/Makefile                                1.45
- sysutils/xenkernel41/distinfo                                1.36
- sysutils/xenkernel41/patches/patch-CVE-2015-2752              1.1
- sysutils/xenkernel41/patches/patch-CVE-2015-2756              1.1
- sysutils/xenkernel42/Makefile                                1.15
- sysutils/xenkernel42/distinfo                                1.13
- sysutils/xenkernel42/patches/patch-CVE-2015-2752              1.1
- sysutils/xenkernel42/patches/patch-CVE-2015-2756              1.1
- sysutils/xenkernel45/Makefile                                1.6
- sysutils/xenkernel45/distinfo                                1.5
- sysutils/xenkernel45/patches/patch-CVE-2015-2752              1.1
- sysutils/xenkernel45/patches/patch-CVE-2015-2756              1.1
- sysutils/xentools41/Makefile                                  1.50
- sysutils/xentools41/distinfo                                  1.38
- sysutils/xentools41/patches/patch-CVE-2015-2752              1.1
- sysutils/xentools41/patches/patch-CVE-2015-2756              1.1
- sysutils/xentools42/Makefile                                  1.27
- sysutils/xentools42/distinfo                                  1.16
- sysutils/xentools42/patches/patch-CVE-2015-2752              1.1
- sysutils/xentools42/patches/patch-CVE-2015-2756              1.1
- sysutils/xentools45/Makefile                                  1.6
- sysutils/xentools45/distinfo                                  1.6
- sysutils/xentools45/patches/patch-CVE-2015-2752              1.1
- sysutils/xentools45/patches/patch-CVE-2015-2756              1.1

---
  Module Name: pkgsrc
  Committed By: spz
  Date: Sun Apr 19 13:13:21 UTC 2015

  Modified Files:
  pkgsrc/sysutils/xenkernel41: Makefile distinfo
  pkgsrc/sysutils/xenkernel42: Makefile distinfo
  pkgsrc/sysutils/xenkernel45: Makefile distinfo
  pkgsrc/sysutils/xentools41: Makefile distinfo
  pkgsrc/sysutils/xentools42: Makefile distinfo
  pkgsrc/sysutils/xentools45: Makefile distinfo
  Added Files:
  pkgsrc/sysutils/xenkernel41/patches: patch-CVE-2015-2752
      patch-CVE-2015-2756
  pkgsrc/sysutils/xenkernel42/patches: patch-CVE-2015-2752
      patch-CVE-2015-2756
  pkgsrc/sysutils/xenkernel45/patches: patch-CVE-2015-2752
      patch-CVE-2015-2756
  pkgsrc/sysutils/xentools41/patches: patch-CVE-2015-2752
      patch-CVE-2015-2756
  pkgsrc/sysutils/xentools42/patches: patch-CVE-2015-2752
      patch-CVE-2015-2756
  pkgsrc/sysutils/xentools45/patches: patch-CVE-2015-2752
      patch-CVE-2015-2756

  Log Message:
  apply fixes from upstream for
  XSA-125 Long latency MMIO mapping operations are not preemptible
  XSA-126 Unmediated PCI command register access in qemu

(tron)

Pullup ticket #4697 - requested by spz
x11/xlockmore: security update
x11/xlockmore-lite: security update

Revisions pulled up:
- x11/xlockmore-lite/Makefile                                  1.21
- x11/xlockmore/Makefile                                        1.119
- x11/xlockmore/Makefile.common                                1.73
- x11/xlockmore/distinfo                                        1.53
- x11/xlockmore/patches/patch-aa                                1.7
- x11/xlockmore/patches/patch-ab                                1.25
- x11/xlockmore/patches/patch-ac                                1.13
- x11/xlockmore/patches/patch-ad                                1.9
- x11/xlockmore/patches/patch-ae                                1.8
- x11/xlockmore/patches/patch-config.cygport                    1.1

---
  Module Name: pkgsrc
  Committed By: spz
  Date: Wed Apr 15 07:31:15 UTC 2015

  Modified Files:
  pkgsrc/x11/xlockmore: Makefile Makefile.common distinfo
  pkgsrc/x11/xlockmore-lite: Makefile
  pkgsrc/x11/xlockmore/patches: patch-aa patch-ab patch-ac patch-ad
      patch-ae
  Added Files:
  pkgsrc/x11/xlockmore/patches: patch-config.cygport

  Log Message:
  Update of xlockmore and xlockmore-lite to the latest version.

  Upstream changelog:

  xlockmore-5.46 as of 10 March 2015, the maintained version of xlock

  5.46
    xmb fonts messing up password screen since ascent was not being set
      correctly.  Also fixed xjack mode.
    Updated language use in xlock.c, now removed to .h files.  If I made some
      bad assumption about the language... let me know.
    BSD install fixes (DESTDIR, FreeType2, fortune) and switching to xz.
    Cygwin install fixes (cygport, xlockrc).

  5.45
    A bad value of fnt for pyro2 could cause to give an X error.  Fixed to
      not free memory we are using and have a better default.

  5.44
    Fix for make_one_sided_pentomino was using 18 as basis for 2 sided
      pentominoes in loop.  Found out about it by gcc 4.8.1 warning: iteration
      12u invokes undefined behavior [-Waggressive-loop-optimizations].
    dilemma, voters, wator size fix for neighbors = 6.
    Proper crediting of Nathan Thompson for discovering High Life part of
      life mode.
    A few new gliders and patterns for life and life3d.
    Updated win32 to use updated cygwin.  Currently links are only satisfied
      using cygwin using 32-bit installation.  For me, cross compiling using
      64-bit gave undefined references in visgl.c and xlock.c.
    magick.c update for version 7, thanks to Jouk Jansen.

  5.43
    life3d: added new rule B58S58, which has a replicator.
    -e added to shell in Makefile.in, told it should now stop on build errors,
      rather than trying to continue which could be confusing.  Thanks to
      Glyn Kennington <glyn AT potatojunkie.co.uk>.
    If crypt fails (which is possible with glibc2.17) it could cause xlock to
      get a signal 11.  There is now protection around the strcmp() to prevent
      NULL poiter dereferences thanks to Mancha <mancha1 AT hush.com>.

  5.42
    USE_MB now defined by default. If this is not defined it can only
      correctly display ASCII.  There could be font issues which may
      lead one to want only ASCII (if so use configure --disable_mb).
      Also looked into Xutf8 but did not see any benefit this had over Xmb.
      Password window and the following modes: bomb,dclock,image,marquee,nose
      are affected.  Fixed a few problems with marquee for width, height, and
      if a character needs more than 2 bytes.
    In Cygwin a seemingly unrelated fix where USE_MB gets defined in xlock.h.
      An #undef was added in xlock.h, if USE_MB gets defined in an internal
      include file.
    -lcrypt added to modes/Makefile.in.  This is needed on my Linux Ubuntu
      and seems ok on Cygwin where it was not needed.  (If this causes a
      problem remove from modes/Makefile).  Also tested on FreeBSD.
    cppcheck fixes
      passwd.c: passcpy could be freed twice
      iostuff.c: could have left file open
      ras.c: leak if mem fails
      resource.c: check if malloc succeeded was in wrong spot
      sound.c: ESound, a redundant check
      memcheck.c: continuing after memory failure
      dclock: problem with parens in ifdefs
      rain: leak fix ( :) )
      t3d: uninitialized time1 (unlikely but fixed anyway)
      molecule: leak if mem fails
      text3d: leak if nothing to draw

(tron)

Pullup ticket #4696 - requested by spz
textproc/libxml2: security patch

Revisions pulled up:
- textproc/libxml2/Makefile                                    1.137
- textproc/libxml2/distinfo                                    1.106
- textproc/libxml2/patches/patch-aa                            1.28
- textproc/libxml2/patches/patch-ab                            1.28
- textproc/libxml2/patches/patch-ac                            1.8
- textproc/libxml2/patches/patch-ad                            1.18
- textproc/libxml2/patches/patch-ae                            1.14
- textproc/libxml2/patches/patch-ag                            1.12
- textproc/libxml2/patches/patch-aj                            1.4
- textproc/libxml2/patches/patch-buf.c                          1.1
- textproc/libxml2/patches/patch-include_libxml_tree.h          1.1
- textproc/libxml2/patches/patch-threads.c                      1.5
- textproc/libxml2/patches/patch-xmlreader.c                    1.1

---
  Module Name: pkgsrc
  Committed By: spz
  Date: Fri Apr 24 11:32:29 UTC 2015

  Modified Files:
  pkgsrc/textproc/libxml2: Makefile distinfo
  pkgsrc/textproc/libxml2/patches: patch-aa patch-ab patch-ac patch-ad
      patch-ae patch-ag patch-aj patch-threads.c
  Added Files:
  pkgsrc/textproc/libxml2/patches: patch-buf.c
      patch-include_libxml_tree.h patch-xmlreader.c

  Log Message:
  patch for CVE-2015-1819 Enforce the reader to run in constant memory
  from https://git.gnome.org/browse/libxml2/commit/?id=213f1fe0d76d30eaed6e5853057defc43e6df2c9

  +general patch refresh

(tron)

+ qt4-creator-3.4.0, stellarium-0.13.3.

(wiz)

Updated emulators/qemu to 2.3.0

(ryoon)

Update to 2.3.0

Changelog:
* Support for 32-bit KVM guests on 64-bit ARM hosts
* Support for running KVM under valgrind
* New IvyBridge CPU model for x86 guests
* Xen: support for ioreq-server API
* New 5KEc and 5KEf MIPS64r2, and M14K and M14Kc MIPS32r2
  microMIPS CPU models for MIPS guests
* Basic support for transactional memory extentions in PowerPC guests
* Improved VGA support for little-endian PPC/pSeries guests
* PCI bus support for s390x guests
* Support for automatic guest device unplug when passthrough devices
  are unbound from VFIO host driver
* Improved UI performance/support for GTK+/VNC/SDL/Spice, and VNC
  support for multiseat
* Performance improvements for virtio-blk emulation: asynchronous SCSI
  request handling, and disk read merging.
* QEMU Guest Agent: now also supports file operations in Windows guests,
  can be used to enable/disable memory blocks in linux guests in
  support for memory hotplug.
* Migration can now include a JSON description of migration stream to aid
  in identifying incompatibilities betweens guests/hosts.
* And lots more...

(ryoon)

pullup 4706

(spz)

Pullup ticket #4706 - requested by roy
net/dnsmasq: security patch

Revisions pulled up:
- net/dnsmasq/Makefile                                          1.28,1.27
- net/dnsmasq/distinfo                                          1.26,1.25
- net/dnsmasq/files/dnsmasq.sh                                  1.3
- net/dnsmasq/patches/patch-src_bpf.c                          1.4
- net/dnsmasq/patches/patch-src_rfc1035.c                      1.1

-------------------------------------------------------------------
  Module Name:    pkgsrc
  Committed By:  roy
  Date:          Wed Apr 29 12:53:02 UTC 2015

  Modified Files:
          pkgsrc/net/dnsmasq: Makefile distinfo
  Added Files:
          pkgsrc/net/dnsmasq/patches: patch-src_rfc1035.c

  Log Message:
  Add a patch to fix CVE-2015-3294.
  Remove myself as maintainer.

  To generate a diff of this commit:
  cvs rdiff -u -r1.27 -r1.28 pkgsrc/net/dnsmasq/Makefile
  cvs rdiff -u -r1.25 -r1.26 pkgsrc/net/dnsmasq/distinfo
  cvs rdiff -u -r0 -r1.1 pkgsrc/net/dnsmasq/patches/patch-src_rfc1035.c

-------------------------------------------------------------------
  Module Name:    pkgsrc
  Committed By:  abs
  Date:          Wed Apr 22 14:57:22 UTC 2015

  Modified Files:
          pkgsrc/net/dnsmasq: Makefile distinfo
          pkgsrc/net/dnsmasq/files: dnsmasq.sh
          pkgsrc/net/dnsmasq/patches: patch-src_bpf.c

  Log Message:
  Add a 'reload' rc.d option (automatically uses the pid file), which will
  make dnsmasq reload the dhcp file. Add patch comment and mps
  Bump PKGREVISION

  To generate a diff of this commit:
  cvs rdiff -u -r1.26 -r1.27 pkgsrc/net/dnsmasq/Makefile
  cvs rdiff -u -r1.24 -r1.25 pkgsrc/net/dnsmasq/distinfo
  cvs rdiff -u -r1.2 -r1.3 pkgsrc/net/dnsmasq/files/dnsmasq.sh
  cvs rdiff -u -r1.3 -r1.4 pkgsrc/net/dnsmasq/patches/patch-src_bpf.c

(spz)

Update to match editors/emacs24/version.mk.  Fixes EMACS_TYPE=emacs24nox.

(jperkin)

do pull in xproto here

(tnn)

- fix false positive for netstat
- fix compilation mistake

(christos)

add x11/libxshmfence

(tnn)

Rename zopen on Mac OS X to avoid conflict with a function by the same
name defined in stdio.h.

(gson)

pullups 4700,4701,4702

(spz)

Pullup ticket #4702 - requested by manu
devel/gnustep-make: build fix

Revisions pulled up:
- devel/gnustep-make/Makefile                                  1.57
- devel/gnustep-make/distinfo                                  1.28
- devel/gnustep-make/patches/patch-Instance_framework.make      1.2

-------------------------------------------------------------------
  Module Name:    pkgsrc
  Committed By:  manu
  Date:          Sun Apr 19 04:13:26 UTC 2015

  Modified Files:
          pkgsrc/devel/gnustep-make: Makefile distinfo
          pkgsrc/devel/gnustep-make/patches: patch-Instance_framework.make

  Log Message:
  Apply fix from upstream, to workaround a build failure in SOGo
  http://svn.gna.org/viewcvs/gnustep?view=revision&revision=36185

  To generate a diff of this commit:
  cvs rdiff -u -r1.56 -r1.57 pkgsrc/devel/gnustep-make/Makefile
  cvs rdiff -u -r1.27 -r1.28 pkgsrc/devel/gnustep-make/distinfo
  cvs rdiff -u -r1.1 -r1.2 \
      pkgsrc/devel/gnustep-make/patches/patch-Instance_framework.make

(spz)

Updated textproc/p5-Text-Autoformat to 1.669006
Updated textproc/p5-Text-CSV to 1.33
Updated print/p5-LaTeX-Encode to 0.091.6
Updated textproc/p5-Text-CSV_XS to 1.17

(mef)

Pullup ticket #4701 - requested by manu
devel/gnustep-objc: build fix

Revisions pulled up:
- devel/gnustep-objc/Makefile                                  1.27
- devel/gnustep-objc/distinfo                                  1.9
- devel/gnustep-objc/patches/patch-objc_msgSend.x86-32.S        1.1

-------------------------------------------------------------------
  Module Name:    pkgsrc
  Committed By:  manu
  Date:          Sat Apr 18 07:23:18 UTC 2015

  Modified Files:
          pkgsrc/devel/gnustep-objc: Makefile distinfo
  Added Files:
          pkgsrc/devel/gnustep-objc/patches: patch-objc_msgSend.x86-32.S

  Log Message:
  Build fix for missing __i686.get_pc_thunk.bx symbol

  Depending on the compiler, __i686.get_pc_thunk.bx may be missing
  from runtime. Replace it by a localy supplied version.

  To generate a diff of this commit:
  cvs rdiff -u -r1.26 -r1.27 pkgsrc/devel/gnustep-objc/Makefile
  cvs rdiff -u -r1.8 -r1.9 pkgsrc/devel/gnustep-objc/distinfo
  cvs rdiff -u -r0 -r1.1 \
      pkgsrc/devel/gnustep-objc/patches/patch-objc_msgSend.x86-32.S

(spz)

Fix DRI3 build on Linux

(tnn)

Update 1.12 to 1.17
-------------------
1.17    - 2015-04-24, H.Merijn Brand
    - Enable overruling $csv in csv ()
    - Allow encoding to be shortened to enc in csv ()
    - Allow filter to alter content
    - Add say (print with default eol => $\)
    - Allow MS sep=; on first line (RT#100304)

1.16    - 2015-03-02, H.Merijn Brand
    * filter made more useful (access to other fields)

1.15    - 2015-02-11, H.Merijn Brand
    * Remove perl recommendation from META as it breaks cpan clients

1.14    - 2015-02-02, H.Merijn Brand
    * Move to github
    * Add csv (filter => {});
    * Change csv ()'s void context behavior

1.13    - 2015-01-03, H.Merijn Brand
    * Simplify code path for old perl
    * Fix quote_binary (#RT100676)
    * Fix csv () for hashrefs with aliased headers
    * Update copyright to 2015

(mef)

Update 0.06 to 0.091.3
----------------------
2014-03-25  Chris Travers <chris.travers@gmail.com>
        * version 0.091.3
        * Further regexp fixes for 5.19 bracket handling

2014-03-24  Chris Travers <chris.travers@gmail.com>
        * version 0.091.1
        * Fixed Test Cases failing taint mode under Perl 5.12 and lower.

2014-03-23  Chris Travers <chris.travers@gmail.com>
        * version 0.09
        * Added additional test case for some entities to make verify table
        * Merged in LaTeX::Encode::EncodingTable

2012-09-30  Andrew Ford  <a.ford@ford-mason.co.uk>
        * version 0.08
        * change t/11-unmatched-chars.t to use an unmatched ideographic character instead
          of 0x{ffff} as some versions of perl protest about having an illegal Unicode
          character in a string.

2012-09-23  Andrew Ford  <a.ford@ford-mason.co.uk>
        * version 0.07
        * remove use of 'parent' in generated module
        * add handling of unmatched characters
        * add note about Unicode data that does not have the UTF8 flag set

(mef)

Fix DRI3 build on Linux

(tnn)

Import libxshmfence-1.2 from pkgsrc-wip as x11/libxshmfence.
(needed to build DRI3 support in MesaLib and modular-xorg-server on Linux)

This library offers a CPU-based synchronization primitive compatible
with the X SyncFence objects that can be shared between processes
using file descriptor passing.

There are three underlying implementations:

1) On Linux, the library uses futexes

2) On FreeBSD, the library uses umtx

3) On other systems, the library uses posix mutexes and condition
    variables.

(tnn)

Remove USE_CWRAPPERS setting, now handled by _USE_CWRAPPERS logic.

(jperkin)

Introduce _USE_CWRAPPERS logic to ensure consistency, USE_CWRAPPERS could
differ between a package and user setting, resulting in incorrect WRAPPER_*
values being set from wrapper-defs.mk.

(jperkin)

Unbreak on FreeBSD by also ensuring that ERSTART is defined and set to the
expected value for it.

Reviewed by bsiegert@

(sevan)

Updated finance/magento to 1.9.1.0nb1

(fhajny)

Fix CVE-2015-1397, CVE-2015-1398 & CVE-2015-1399 via upstream patches.
Implement way to apply upstream patches using PATCHFILES.
Bump PKGREVISION.

(fhajny)

Update to 1.33
--------------
1.33  Tue Mar  3 14:13:41 2015
        - Fix some reported bugs with handling fields with zeros.
          https://rt.cpan.org/Public/Bug/Display.html?id=93518
          https://rt.cpan.org/Public/Bug/Display.html?id=92509
                                              patched by shlomif
        - catch up Text::CSV_XS 1.02
            * add decode_utf8 attribute by ktat
        - fix warnings: $. is undefined by ktat
        - typo fixes in doc by anirvan, dsteinbrunner and pdl.
        - parse() supports bind_columns()

        * TODO: Updating documents and adding diag_verbose in the next version.
                Catching up CSV_XS 1.04.

(mef)

Update 1.669004 to 1.669006
---------------------------
1.669005  Fri Oct  3 16:56:12 2014
    - Added Lisp-style leading ; as a quoter (thanks Roland)

    - Added 'quoter' option to allow user to specify a regex
      to be used to recognize quoters (thanks Peter)

1.669006  Sat Oct  4 07:32:33 2014
    - Improved 'quoter' option to allow augmenting quoter regex,
      rather than just replacing it (thanks Peter)

(mef)

Use a single egrep and regexp instead of iterating over the list of files to
check and calling grep on each of them.  Idea for using ":ts|" from Joerg.

Reduces the test runtime by around 50%, depending on platform and package.

(jperkin)

Only check executable files for valid interpreters.  By the time this check is
performed all installed files should have their correct mode set, and there are
no INSTALL scripts that currently change file modes under PREFIX after
installation (nor should they ever).

This change considerably speeds up the test.  For example in www/firefox-l10n
the time to run the test reduces from 181 seconds (135 seconds of system time)
down to 10 seconds (6 seconds system time) on my SmartOS build systems.

(jperkin)

Add a patch to fix CVE-2015-3294.
Remove myself as maintainer.

(roy)

Updated devel/p5-B-Utils to 0.26
Updated devel/p5-CPAN-FindDependencies to 2.43
Updated devel/p5-Want to 0.26

(mef)

Update to 0.26
--------------
0.26  Tue 28 Apr 2015 12:31:17 BST
    - There is a new experimental option in bleadperl under which one ought not to use
      the op_sibling field directly. Quoting from perlguts.pod:

        Starting in version 5.21.2, perls built with the experimental
        define C<-DPERL_OP_PARENT> add an extra boolean flag for each op,
        C<op_moresib>.  When not set, this indicates that this is the last op in an
        C<OpSIBLING> chain. This frees up the C<op_sibling> field on the last
        sibling to point back to the parent op. Under this build, that field is
        also renamed C<op_sibparent> to reflect its joint role. The macro
        C<OpSIBLING(o)> wraps this special behaviour, and always returns NULL on
        the last sibling.  With this build the C<op_parent(o)> function can be
        used to find the parent of any op. Thus for forward compatibility, you
        should always use the C<OpSIBLING(o)> macro rather than accessing
        C<op_sibling> directly.

      Thanks to Reini Urban for the patch.

(mef)

Update to 2.43
--------------
2.43    2015-04-26      Make segfault go away on Windows

(mef)

Update to 0.26
--------------
0.26
  - fix tests for perl 5.21.7 (change to nulled COPs) (RT#100508, Father
    Chrysostomos)

(mef)

Updated devel/pcre to 8.37

(mef)

(pkgsrc)
- Drop patch-CVE-2014-8964 (now included)
(upstream)
- Update to 8.37
Version 8.37 28-April-2015
--------------------------
1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
    for those parentheses to be closed with whatever has been captured so far.
    However, it was failing to mark any other groups between the hightest
    capture so far and the currrent group as "unset". Thus, the ovector for
    those groups contained whatever was previously there. An example is the
    pattern /(x)|((*ACCEPT))/ when matched against "abcd".

2.  If an assertion condition was quantified with a minimum of zero (an odd
    thing to do, but it happened), SIGSEGV or other misbehaviour could occur.

3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
    unrecognized modifier, a crash could occur.

4.  An attempt to do global matching in pcretest with a zero-length ovector
    caused a crash.

5.  Fixed a memory leak during matching that could occur for a subpattern
    subroutine call (recursive or otherwise) if the number of captured groups
    that had to be saved was greater than ten.

6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
    string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
    UTF with NO_UTF_CHECK is documented as having an undefined outcome.

7.  A UTF pattern containing a "not" match of a non-ASCII character and a
    subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.

8. When a pattern is compiled, it remembers the highest back reference so that
  when matching, if the ovector is too small, extra memory can be obtained to
  use instead. A conditional subpattern whose condition is a check on a
  capture having happened, such as, for example in the pattern
  /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
  setting the highest backreference number. This mattered only if pcre_exec()
  was called with an ovector that was too small to hold the capture, and there
  was no other kind of back reference (a situation which is probably quite
  rare). The effect of the bug was that the condition was always treated as
  FALSE when the capture could not be consulted, leading to a incorrect
  behaviour by pcre_exec(). This bug has been fixed.

9. A reference to a duplicated named group (either a back reference or a test
  for being set in a conditional) that occurred in a part of the pattern where
  PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
  to be incorrectly calculated, leading to overwriting.

10. A mutually recursive set of back references such as (\2)(\1) caused a
    segfault at study time (while trying to find the minimum matching length).
    The infinite loop is now broken (with the minimum length unset, that is,
    zero).

11. If an assertion that was used as a condition was quantified with a minimum
    of zero, matching went wrong. In particular, if the whole group had
    unlimited repetition and could match an empty string, a segfault was
    likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
    assertions to be quantified, but not if they are being used as conditions,
    so the above pattern is faulted by Perl. PCRE has now been changed so that
    it also rejects such patterns.

12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
    failed to allow the zero-repeat case if pcre2_exec() was called with an
    ovector too small to capture the group.

13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
    Red Hat Product Security:

    (a) A crash if /K and /F were both set with the option to save the compiled
    pattern.

    (b) Another crash if the option to print captured substrings in a callout
    was combined with setting a null ovector, for example \O\C+ as a subject
    string.

14. A pattern such as "((?2){0,1999}())?", which has a group containing a
    forward reference repeated a large (but limited) number of times within a
    repeated outer group that has a zero minimum quantifier, caused incorrect
    code to be compiled, leading to the error "internal error:
    previously-checked referenced subpattern not found" when an incorrect
    memory address was read. This bug was reported as "heap overflow",
    discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
    CVE-2015-2325.

23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
    call within a group that also contained a recursive back reference caused
    incorrect code to be compiled. This bug was reported as "heap overflow",
    discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
    number CVE-2015-2326.

24. Computing the size of the JIT read-only data in advance has been a source
    of various issues, and new ones are still appear unfortunately. To fix
    existing and future issues, size computation is eliminated from the code,
    and replaced by on-demand memory allocation.

25. A pattern such as /(?i)[A-`]/, where characters in the other case are
    adjacent to the end of the range, and the range contained characters with
    more than one other case, caused incorrect behaviour when compiled in UTF
    mode. In that example, the range a-j was left out of the class.

26. Fix JIT compilation of conditional blocks, which assertion
    is converted to (*FAIL). E.g: /(?(?!))/.

27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
    discovered by the LLVM fuzzer.

28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
    when this assertion was used as a condition, for example (?(?!)a|b). In
    pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
    error about an unsupported item.

29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
    possessification code could take exponential time to complete. A recursion
    depth limit of 1000 has been imposed to limit the resources used by this
    optimization.

30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
    such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
    because \S ensures they are all in the class. The code for doing this was
    interacting badly with the code for computing the amount of space needed to
    compile the pattern, leading to a buffer overflow. This bug was discovered
    by the LLVM fuzzer.

31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
    other kinds of group caused stack overflow at compile time. This bug was
    discovered by the LLVM fuzzer.

32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
    between a subroutine call and its quantifier was incorrectly compiled,
    leading to buffer overflow or other errors. This bug was discovered by the
    LLVM fuzzer.

33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
    assertion after (?(. The code was failing to check the character after
    (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
    was discovered by the LLVM fuzzer.

34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
    a fixed maximum following a group that contains a subroutine reference was
    incorrectly compiled and could trigger buffer overflow. This bug was
    discovered by the LLVM fuzzer.

35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
    caused a stack overflow instead of the diagnosis of a non-fixed length
    lookbehind assertion. This bug was discovered by the LLVM fuzzer.

36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
    (e.g. /(?<=\Ka)/) could make pcregrep loop.

37. There was a similar problem to 36 in pcretest for global matches.

38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
    and a subsequent item in the pattern caused a non-match, backtracking over
    the repeated \X did not stop, but carried on past the start of the subject,
    causing reference to random memory and/or a segfault. There were also some
    other cases where backtracking after \C could crash. This set of bugs was
    discovered by the LLVM fuzzer.

39. The function for finding the minimum length of a matching string could take
    a very long time if mutual recursion was present many times in a pattern,
    for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
    method has been implemented. This infelicity was discovered by the LLVM
    fuzzer.

40. Static linking against the PCRE library using the pkg-config module was
    failing on missing pthread symbols.

(mef)

Updated chat/ejabberd to 15.04

(fhajny)

Update chat/ejabberd to 15.04. Add sqlite option.

Upstream changes:
- R16B03-1 is now the minimal required Erlang/OTP version
- SQLite support
- Default db_type can be specified with global default_db option
- Included mod_muc_admin, mod_admin_extra and mod_multicast modules
- Removed ejabberd_http_poll
- Pubsub improvements
- Better RFC compliance
- Several other bugfixes

(fhajny)

Fix PLIST for the erlang-hipe option. Fixes default build on NetBSD.

(fhajny)

Updated time/ocaml-calendar to 2.04

(jaapb)