oweals/openssl.git
8 years agoDon't test quite so many of them.
David Benjamin [Wed, 10 Aug 2016 14:45:49 +0000 (10:45 -0400)]
Don't test quite so many of them.

Avoid making the CI blow up.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 5cf6d7c51f16fd78de7921dc441e24897c8b3cc6)

8 years agoTest CBC mode padding.
David Benjamin [Wed, 10 Aug 2016 04:45:51 +0000 (00:45 -0400)]
Test CBC mode padding.

This is a regression test for
https://github.com/openssl/openssl/pull/1431. It tests a
maximally-padded record with each possible invalid offset.

This required fixing a bug in Message.pm where the client sending a
fatal alert followed by close_notify was still treated as success.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 8523288e6d667f052bda092e01ab17986782fede)

8 years agoFix some mem leaks in sslapitest
Matt Caswell [Mon, 26 Sep 2016 11:04:23 +0000 (12:04 +0100)]
Fix some mem leaks in sslapitest

A mem leak could occur on an error path. Also the mempacket BIO_METHOD
needs to be cleaned up, because of the newly added DTLS test.

Also fixed a double semi-colon in ssltestlib.c

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit fa454945cf2855fed452ff9bdb1876096bc07beb)

8 years agoPrepare for 1.1.0c-dev
Matt Caswell [Mon, 26 Sep 2016 09:46:58 +0000 (10:46 +0100)]
Prepare for 1.1.0c-dev

Reviewed-by: Richard Levitte <levitte@openssl.org>
8 years agoPrepare for 1.1.0b release OpenSSL_1_1_0b
Matt Caswell [Mon, 26 Sep 2016 09:46:03 +0000 (10:46 +0100)]
Prepare for 1.1.0b release

Reviewed-by: Richard Levitte <levitte@openssl.org>
8 years agoUpdates CHANGES and NEWS for new release
Matt Caswell [Mon, 26 Sep 2016 08:43:45 +0000 (09:43 +0100)]
Updates CHANGES and NEWS for new release

Reviewed-by: Richard Levitte <levitte@openssl.org>
8 years agoAdd to fuzz corpora for CVE-2016-6309
Robert Swiecki [Sun, 25 Sep 2016 15:35:56 +0000 (16:35 +0100)]
Add to fuzz corpora for CVE-2016-6309

Reviewed-by: Emilia Käsper <emilia@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 44f206aa9dfd4f226f17d9093732dbece5300aa6)

8 years agoFix Use After Free for large message sizes
Matt Caswell [Fri, 23 Sep 2016 15:58:11 +0000 (16:58 +0100)]
Fix Use After Free for large message sizes

The buffer to receive messages is initialised to 16k. If a message is
received that is larger than that then the buffer is "realloc'd". This can
cause the location of the underlying buffer to change. Anything that is
referring to the old location will be referring to free'd data. In the
recent commit c1ef7c97 (master) and 4b390b6c (1.1.0) the point in the code
where the message buffer is grown was changed. However s->init_msg was not
updated to point at the new location.

CVE-2016-6309

Reviewed-by: Emilia Käsper <emilia@openssl.org>
(cherry picked from commit 0d698f6696e114a6e47f8b75ff88ec81f9e30175)

8 years agoAdd a test for large messages
Matt Caswell [Fri, 23 Sep 2016 14:37:13 +0000 (15:37 +0100)]
Add a test for large messages

Ensure that we send a large message during the test suite.

Reviewed-by: Emilia Käsper <emilia@openssl.org>
(cherry picked from commit 84d5549e692e63a16fa1b11603e4098fc31746e9)

8 years agoPrepare for 1.1.0b-dev
Matt Caswell [Thu, 22 Sep 2016 10:15:54 +0000 (11:15 +0100)]
Prepare for 1.1.0b-dev

Reviewed-by: Richard Levitte <levitte@openssl.org>
8 years agoPrepare for 1.1.0a release OpenSSL_1_1_0a
Matt Caswell [Thu, 22 Sep 2016 10:14:50 +0000 (11:14 +0100)]
Prepare for 1.1.0a release

Reviewed-by: Richard Levitte <levitte@openssl.org>
8 years agoUpdates CHANGES and NEWS for new release
Matt Caswell [Wed, 21 Sep 2016 20:59:49 +0000 (21:59 +0100)]
Updates CHANGES and NEWS for new release

Reviewed-by: Richard Levitte <levitte@openssl.org>
8 years agoAvoid KCI attack for GOST
Dmitry Belyavsky [Mon, 19 Sep 2016 14:53:35 +0000 (15:53 +0100)]
Avoid KCI attack for GOST

Russian GOST ciphersuites are vulnerable to the KCI attack because they use
long-term keys to establish the connection when ssl client authorization is
on. This change brings the GOST implementation into line with the latest
specs in order to avoid the attack. It should not break backwards
compatibility.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
8 years agoFix a hang with SSL_peek()
Matt Caswell [Sat, 10 Sep 2016 20:24:40 +0000 (21:24 +0100)]
Fix a hang with SSL_peek()

If while calling SSL_peek() we read an empty record then we go into an
infinite loop, continually trying to read data from the empty record and
never making any progress. This could be exploited by a malicious peer in
a Denial Of Service attack.

CVE-2016-6305

GitHub Issue #1563

Reviewed-by: Rich Salz <rsalz@openssl.org>
8 years agoFix a mem leak in NPN handling
Matt Caswell [Fri, 9 Sep 2016 09:53:39 +0000 (10:53 +0100)]
Fix a mem leak in NPN handling

If a server sent multiple NPN extensions in a single ClientHello then a
mem leak can occur. This will only happen where the client has requested
NPN in the first place. It does not occur during renegotiation. Therefore
the maximum that could be leaked in a single connection with a malicious
server is 64k (the maximum size of the ServerHello extensions section). As
this is client side, only occurs if NPN has been requested and does not
occur during renegotiation this is unlikely to be exploitable.

Issue reported by Shi Lei.

Reviewed-by: Rich Salz <rsalz@openssl.org>
8 years agoAdd some more OCSP testing
Matt Caswell [Tue, 13 Sep 2016 16:02:03 +0000 (17:02 +0100)]
Add some more OCSP testing

Test that the OCSP callbacks work as expected.

Reviewed-by: Rich Salz <rsalz@openssl.org>
8 years agoAdd OCSP_RESPID_match()
Matt Caswell [Tue, 13 Sep 2016 22:26:53 +0000 (23:26 +0100)]
Add OCSP_RESPID_match()

Add a function for testing whether a given OCSP_RESPID matches with a
certificate.

Reviewed-by: Rich Salz <rsalz@openssl.org>
8 years agoAdd the ability to set OCSP_RESPID fields
Matt Caswell [Mon, 12 Sep 2016 16:39:55 +0000 (17:39 +0100)]
Add the ability to set OCSP_RESPID fields

OCSP_RESPID was made opaque in 1.1.0, but no accessors were provided for
setting the name/key value for the OCSP_RESPID.

Reviewed-by: Rich Salz <rsalz@openssl.org>
8 years agoFix OCSP Status Request extension unbounded memory growth
Matt Caswell [Fri, 9 Sep 2016 09:08:45 +0000 (10:08 +0100)]
Fix OCSP Status Request extension unbounded memory growth

A malicious client can send an excessively large OCSP Status Request
extension. If that client continually requests renegotiation,
sending a large OCSP Status Request extension each time, then there will
be unbounded memory growth on the server. This will eventually lead to a
Denial Of Service attack through memory exhaustion. Servers with a
default configuration are vulnerable even if they do not support OCSP.
Builds using the "no-ocsp" build time option are not affected.

I have also checked other extensions to see if they suffer from a similar
problem but I could not find any other issues.

CVE-2016-6304

Issue reported by Shi Lei.

Reviewed-by: Rich Salz <rsalz@openssl.org>
8 years agoFix error message typo, wrong function code
Richard Levitte [Thu, 22 Sep 2016 08:15:02 +0000 (10:15 +0200)]
Fix error message typo, wrong function code

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit a449b47c7d8e20efc8cc524ed695a060b11ef889)

8 years agotest/x509aux.c: Fix argv loop
Richard Levitte [Wed, 21 Sep 2016 12:44:42 +0000 (14:44 +0200)]
test/x509aux.c: Fix argv loop

There are cases when argc is more trustable than proper argv termination.
Since we trust argc in all other test programs, we might as well treat it
the same way in this program.

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 780bbb96bf514f0b4013e9c5725614ba5153c497)

8 years agoCHANGES: mention removal of solaris-x86-cc.
Andy Polyakov [Wed, 21 Sep 2016 20:05:12 +0000 (22:05 +0200)]
CHANGES: mention removal of solaris-x86-cc.

Reviewed-by: Rich Salz <rsalz@openssl.org>
8 years agoExcessive allocation of memory in dtls1_preprocess_fragment()
Matt Caswell [Mon, 19 Sep 2016 10:57:07 +0000 (11:57 +0100)]
Excessive allocation of memory in dtls1_preprocess_fragment()

This issue is very similar to CVE-2016-6307 described in the previous
commit. The underlying defect is different but the security analysis and
impacts are the same except that it impacts DTLS.

A DTLS message includes 3 bytes for its length in the header for the
message.
This would allow for messages up to 16Mb in length. Messages of this length
are excessive and OpenSSL includes a check to ensure that a peer is sending
reasonably sized messages in order to avoid too much memory being consumed
to service a connection. A flaw in the logic of version 1.1.0 means that
memory for the message is allocated too early, prior to the excessive
message length check. Due to way memory is allocated in OpenSSL this could
mean an attacker could force up to 21Mb to be allocated to service a
connection. This could lead to a Denial of Service through memory
exhaustion. However, the excessive message length check still takes place,
and this would cause the connection to immediately fail. Assuming that the
application calls SSL_free() on the failed conneciton in a timely manner
then the 21Mb of allocated memory will then be immediately freed again.
Therefore the excessive memory allocation will be transitory in nature.
This then means that there is only a security impact if:

1) The application does not call SSL_free() in a timely manner in the
event that the connection fails
or
2) The application is working in a constrained environment where there
is very little free memory
or
3) The attacker initiates multiple connection attempts such that there
are multiple connections in a state where memory has been allocated for
the connection; SSL_free() has not yet been called; and there is
insufficient memory to service the multiple requests.

Except in the instance of (1) above any Denial Of Service is likely to
be transitory because as soon as the connection fails the memory is
subsequently freed again in the SSL_free() call. However there is an
increased risk during this period of application crashes due to the lack
of memory - which would then mean a more serious Denial of Service.

This issue does not affect TLS users.

Issue was reported by Shi Lei (Gear Team, Qihoo 360 Inc.).

CVE-2016-6308

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 48c054fec3506417b2598837b8062aae7114c200)

8 years agoExcessive allocation of memory in tls_get_message_header()
Matt Caswell [Mon, 19 Sep 2016 10:39:21 +0000 (11:39 +0100)]
Excessive allocation of memory in tls_get_message_header()

A TLS message includes 3 bytes for its length in the header for the message.
This would allow for messages up to 16Mb in length. Messages of this length
are excessive and OpenSSL includes a check to ensure that a peer is sending
reasonably sized messages in order to avoid too much memory being consumed
to service a connection. A flaw in the logic of version 1.1.0 means that
memory for the message is allocated too early, prior to the excessive
message length check. Due to way memory is allocated in OpenSSL this could
mean an attacker could force up to 21Mb to be allocated to service a
connection. This could lead to a Denial of Service through memory
exhaustion. However, the excessive message length check still takes place,
and this would cause the connection to immediately fail. Assuming that the
application calls SSL_free() on the failed conneciton in a timely manner
then the 21Mb of allocated memory will then be immediately freed again.
Therefore the excessive memory allocation will be transitory in nature.
This then means that there is only a security impact if:

1) The application does not call SSL_free() in a timely manner in the
event that the connection fails
or
2) The application is working in a constrained environment where there
is very little free memory
or
3) The attacker initiates multiple connection attempts such that there
are multiple connections in a state where memory has been allocated for
the connection; SSL_free() has not yet been called; and there is
insufficient memory to service the multiple requests.

Except in the instance of (1) above any Denial Of Service is likely to
be transitory because as soon as the connection fails the memory is
subsequently freed again in the SSL_free() call. However there is an
increased risk during this period of application crashes due to the lack
of memory - which would then mean a more serious Denial of Service.

This issue does not affect DTLS users.

Issue was reported by Shi Lei (Gear Team, Qihoo 360 Inc.).

CVE-2016-6307

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit c1ef7c971d0bbf117c3c80f65b5875e2e7b024b1)

8 years agoConfigure: clarify and refine -static.
Andy Polyakov [Thu, 1 Sep 2016 19:36:13 +0000 (21:36 +0200)]
Configure: clarify and refine -static.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 047d97afd97520eae268f6d8a36fbf9a0239a994)

8 years agoDon't leak on an OPENSSL_realloc() failure
Matt Caswell [Wed, 21 Sep 2016 14:49:28 +0000 (15:49 +0100)]
Don't leak on an OPENSSL_realloc() failure

If OPENSSL_sk_insert() calls OPENSSL_realloc() and it fails, it was leaking
the originally allocated memory.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 41bff723c6784cc846054a4fd4add6dbec8c2c64)

8 years agoDon't allow too many consecutive warning alerts
Matt Caswell [Wed, 21 Sep 2016 13:07:31 +0000 (14:07 +0100)]
Don't allow too many consecutive warning alerts

Certain warning alerts are ignored if they are received. This can mean that
no progress will be made if one peer continually sends those warning alerts.
Implement a count so that we abort the connection if we receive too many.

Issue reported by Shi Lei.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit af58be768ebb690f78530f796e92b8ae5c9a4401)

8 years agorand/randfile.c: treat empty string in RAND_file_name as error.
Andy Polyakov [Tue, 20 Sep 2016 15:08:03 +0000 (17:08 +0200)]
rand/randfile.c: treat empty string in RAND_file_name as error.

Suggested in GH#1589.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 7dc0ad4d6dca81a003be7fa1fbd58a55f4be8646)

8 years agorand/randfile.c: rationalize __OpenBSD__ code path.
Andy Polyakov [Tue, 20 Sep 2016 15:06:58 +0000 (17:06 +0200)]
rand/randfile.c: rationalize __OpenBSD__ code path.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit ba8fa4e53a35a0c46d1e0e81a4d270a026fac2b9)

8 years agorand/randfile.c: restore fallback to $HOME for non-setuid programs.
Andy Polyakov [Tue, 20 Sep 2016 14:59:32 +0000 (16:59 +0200)]
rand/randfile.c: restore fallback to $HOME for non-setuid programs.

Reported in GH#1589, but solution is different from suggested.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 799c1293fcf412db64dcc8a09a6b11cc755914dc)

8 years agoDcoument -alpn flag
Rich Salz [Wed, 21 Sep 2016 14:59:15 +0000 (10:59 -0400)]
Dcoument -alpn flag

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 776e15f9393a9e3083bec60a8da376ce2fe1e97e)

8 years agoGH1555: Don't bump size on realloc failure
Rich Salz [Fri, 9 Sep 2016 14:52:59 +0000 (10:52 -0400)]
GH1555: Don't bump size on realloc failure

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 6fcace45bda108ad4d3f95261494dd479720d92c)

8 years agoUse switch instead of multiple ifs
Alessandro Ghedini [Tue, 13 Sep 2016 23:51:02 +0000 (00:51 +0100)]
Use switch instead of multiple ifs

Makes the logic a little bit clearer.

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1571)
(cherry picked from commit 4f8a5f4da94bb70e06cf709beb1ddfa57a218c3d)

8 years agoFix a missing NULL check in dsa_builtin_paramgen
Matt Caswell [Wed, 14 Sep 2016 12:27:59 +0000 (13:27 +0100)]
Fix a missing NULL check in dsa_builtin_paramgen

We should check the last BN_CTX_get() call to ensure that it isn't NULL
before we try and use any of the allocated BIGNUMs.

Issue reported by Shi Lei.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 1ff7425d6130380bb00d3e64739633a4b21b11a3)

8 years agoVMS: add [.util]shlib_wrap.exe and its build instructions
Richard Levitte [Tue, 20 Sep 2016 23:47:06 +0000 (01:47 +0200)]
VMS: add [.util]shlib_wrap.exe and its build instructions

This is a program for VMS that corresponds to util/shlib_wrap.sh.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit f3ff481f318b10a223d6157bde9645e1797487c5)

8 years agodescrip.mms.tmpl: in obj2bin, make sure an empty @deps means no empty lines
Richard Levitte [Tue, 20 Sep 2016 23:49:53 +0000 (01:49 +0200)]
descrip.mms.tmpl: in obj2bin, make sure an empty @deps means no empty lines

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 47852e6ae763a40ddc3538c76a02be06fd0595a5)

8 years agoutil/dofile.pl: report if a template couldn't be loaded
Richard Levitte [Tue, 20 Sep 2016 23:49:04 +0000 (01:49 +0200)]
util/dofile.pl: report if a template couldn't be loaded

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit f6be8315cbdb4173008b5917d2b0fc80bb0bf06b)

8 years agoRT3669: dgst can only sign/verify one file.
Rich Salz [Fri, 2 Sep 2016 21:34:31 +0000 (17:34 -0400)]
RT3669: dgst can only sign/verify one file.

Check arg count and print an error message.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 13a461831ab85ec9f55806ebf2df5ffb7f3f357a)

8 years agoapps/apps.c: include sys/socket.h to declare recv()
Richard Levitte [Tue, 20 Sep 2016 16:43:24 +0000 (18:43 +0200)]
apps/apps.c: include sys/socket.h to declare recv()

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit a19228b7f4fc6fcb49713455b3caedbc24fb0b01)

8 years agoDocumentation fixup; no more ECDHParameters
Richard Levitte [Sun, 18 Sep 2016 21:52:30 +0000 (23:52 +0200)]
Documentation fixup; no more ECDHParameters

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 6e836806add140fa9e56d1226d7514fdfa2c803a)

8 years agoCrude VMS build files for demos/bio/
Richard Levitte [Thu, 8 Sep 2016 23:53:38 +0000 (01:53 +0200)]
Crude VMS build files for demos/bio/

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 8c88c88b5ad43bbdf1f6e7602b6ac56c9031b153)

8 years agoCrude Makefile for demos/bio/
Richard Levitte [Thu, 8 Sep 2016 23:01:30 +0000 (01:01 +0200)]
Crude Makefile for demos/bio/

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit f4566dff06d2539dd01f54c817e2b952b64452e4)

8 years agoDon't ignore Makefiles in demos/
Richard Levitte [Thu, 8 Sep 2016 23:01:15 +0000 (01:01 +0200)]
Don't ignore Makefiles in demos/

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit ccf11f7ee49a0825caa407ed481c21b4b5933023)

8 years agoFixup BIO demos for OpenSSL 1.1.x
Richard Levitte [Fri, 9 Sep 2016 07:32:00 +0000 (09:32 +0200)]
Fixup BIO demos for OpenSSL 1.1.x

Call SSL_CTX_new() before doing any configuration.
(or call OPENSSL_ssl_init())

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 54d72ef0f019db383f8e98342b9b77c5da8541e5)

8 years agoFixup BIO demos for OpenSSL 1.1.x
Richard Levitte [Thu, 8 Sep 2016 22:59:00 +0000 (00:59 +0200)]
Fixup BIO demos for OpenSSL 1.1.x

'ECDHParameters = Automatic' isn't accepted.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 3656ea1c2217f38d44f3f34253e7c16f1b40ba77)

8 years agoFixup BIO demos for OpenSSL 1.1.x
Richard Levitte [Thu, 8 Sep 2016 22:58:21 +0000 (00:58 +0200)]
Fixup BIO demos for OpenSSL 1.1.x

Note: server-cmod doesn't seem to do things right...  from loading
cmod.cnf, it tries to load libssl_conf.so.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 16c6deed2c42d4cf4a7676a32b718d6a867f482a)

8 years agoRemove an option related to a deprecated flag
FdaSilvaYY [Sat, 17 Sep 2016 19:29:48 +0000 (21:29 +0200)]
Remove an option related to a deprecated flag

CMS_NOOLDMIMETYPE and PKCS7_NOOLDMIMETYPE  are unused in pkcs7/cms code.

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1585)
(cherry picked from commit 28aef3d9558dc2e11ba56576b3a4d3faaef8a9d3)

8 years agoFix various missing option help messages ...
FdaSilvaYY [Sun, 31 Jul 2016 17:02:50 +0000 (19:02 +0200)]
Fix various missing option help messages ...

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1585)
(cherry picked from commit 12d56b2992ebd61e1b30c99ca1898dde42345cf7)

8 years agoConvert num_alloc to a size_t in stack.c and tweak style
Matt Caswell [Thu, 8 Sep 2016 10:06:29 +0000 (11:06 +0100)]
Convert num_alloc to a size_t in stack.c and tweak style

We were casting num_alloc to size_t in lots of places, or just using it in
a context where size_t makes more sense - so convert it. This simplifies
the code a bit.

Also tweak the style in stack.c a bit following on from the previous
commit

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 9205ebeb8e448b2d6948b9e5d78ecf309c0ed33c)

8 years agoPrevent overflows in stack API
Guido Vranken [Thu, 8 Sep 2016 09:43:37 +0000 (10:43 +0100)]
Prevent overflows in stack API

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 9731a9ce7d0f404d21ed418f9bc983b174e130cb)

8 years agoUpdate INSTALL about no-md5 removal
FdaSilvaYY [Mon, 12 Sep 2016 20:42:37 +0000 (22:42 +0200)]
Update INSTALL about no-md5 removal

no-rsa is no longer an option since 7ec8de1

Fix a typo about poly1305

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1582)
(cherry picked from commit e2562bbbe1e1c68ec5a3e02c1f151fd6149ee2ae)

8 years agoFix no-ocsp
Richard Levitte [Mon, 19 Sep 2016 13:08:58 +0000 (15:08 +0200)]
Fix no-ocsp

Some compilers complain about unused variables, and some tests do
not run well without OCSP.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 057c676afb2a69b7d2bb9cf6e895e720a2a62b78)

8 years agoHave the configuration options 'no-err' and 'no-async' work again
Richard Levitte [Sat, 17 Sep 2016 19:57:29 +0000 (21:57 +0200)]
Have the configuration options 'no-err' and 'no-async' work again

In an earlier attempt to simplify the processing of disabled options,
'no-err' and 'no-async' stopped working properly.  'err' and 'async'
are directories under 'crypto/', but they are special insofar that
they can't be simply skipped, like all the algorithm directories can,
so they need special treatment among the disablable things.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit 66fe388aa410820d80ab1d99730b64b1b56a89d4)

8 years agoSimplify configudata.pm dependency generation
Richard Levitte [Sat, 17 Sep 2016 18:50:56 +0000 (20:50 +0200)]
Simplify configudata.pm dependency generation

While we're at it, correct the fault in windows-makefile.tmpl

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit 41240e68d745972d56dd037caf439d9c5032813f)

8 years agoRemove an useless definition.
FdaSilvaYY [Thu, 15 Sep 2016 18:33:18 +0000 (20:33 +0200)]
Remove an useless definition.

APP_INFO is currently a field of MEM struct.

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1583)
(cherry picked from commit 76dc356384b3d8f93f04be05f320a878e4bced52)

8 years agoSimplify the dependencies for remaking configdata.pm
Richard Levitte [Thu, 15 Sep 2016 21:56:03 +0000 (23:56 +0200)]
Simplify the dependencies for remaking configdata.pm

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit c5ad3076a6c076ac9f8449c393dd14c16b737560)

8 years agoRearrange the storage of build file template names in %config
Richard Levitte [Thu, 15 Sep 2016 21:55:24 +0000 (23:55 +0200)]
Rearrange the storage of build file template names in %config

They are now relative paths as well

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit 1967a42eb5192efc552287f78850420082caba7c)

8 years agoUse the registered conf file names as dependencies to build configdata.pm
Richard Levitte [Wed, 14 Sep 2016 20:22:20 +0000 (22:22 +0200)]
Use the registered conf file names as dependencies to build configdata.pm

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit dc660c70b52c552aa4eb9659963bfbe6ab18bec0)

8 years agoRegister the name of the config file each config target was found in
Richard Levitte [Wed, 14 Sep 2016 20:21:41 +0000 (22:21 +0200)]
Register the name of the config file each config target was found in

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit ee9b0bbb4c66812da18939ea96c4625cb001f3bd)

8 years agoRevert "Abort on unrecognised warning alerts"
Matt Caswell [Thu, 15 Sep 2016 19:43:16 +0000 (20:43 +0100)]
Revert "Abort on unrecognised warning alerts"

This reverts commit 77a6be4dfc2ecf406c2559a99bea51317ce0f533.

There were some unexpected side effects to this commit, e.g. in SSLv3 a
warning alert gets sent "no_certificate" if a client does not send a
Certificate during Client Auth. With the above commit this causes the
connection to abort, which is incorrect. There may be some other edge cases
like this so we need to have a rethink on this.

Reviewed-by: Tim Hudson <tjh@openssl.org>
8 years agoFinally, make sure vms_term_sock.c is built on VMS
Richard Levitte [Wed, 14 Sep 2016 21:40:27 +0000 (23:40 +0200)]
Finally, make sure vms_term_sock.c is built on VMS

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit d602c2b680aefd3e0d00d090783ef5f912caf06a)

8 years agoRefactor to avoid unnecessary preprocessor logic
Richard Levitte [Thu, 15 Sep 2016 09:20:18 +0000 (11:20 +0200)]
Refactor to avoid unnecessary preprocessor logic

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 51e5133d551b4c132f72fc2ff5bbe076f5a3e052)

8 years agoReformat to fit OpenSSL source code standards
Richard Levitte [Wed, 14 Sep 2016 18:54:30 +0000 (20:54 +0200)]
Reformat to fit OpenSSL source code standards

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit c7bdb6a31ff0fcae66b451d3f80a684ad77f4966)

8 years agoAdd copyright and license on apps/vms_term_sock.[ch]
Richard Levitte [Wed, 14 Sep 2016 18:52:03 +0000 (20:52 +0200)]
Add copyright and license on apps/vms_term_sock.[ch]

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 0d0723e813870c2e7a009ec47e933668a1548531)

8 years agoVSI submission: redirect terminal input through socket
Richard Levitte [Wed, 3 Aug 2016 19:16:43 +0000 (21:16 +0200)]
VSI submission: redirect terminal input through socket

This is needed, because on VMS, select() can only be used on sockets.  being
able to use select() on all kinds of file descriptors is unique to Unix.

So, the solution for VMS is to create a layer that translates input from
standard input to socket communication.

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 0d3b65832c6fa94c1d1cfa2f99335f323e3227c1)

8 years agoMake reference to other manpage more explicit
Rich Salz [Wed, 14 Sep 2016 15:13:57 +0000 (11:13 -0400)]
Make reference to other manpage more explicit

Where -curves, etc., are defined: SSL_CONF_cmd

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit ddb30bde105a50d6a79a9662eeaf27e8cb29c606)

8 years agoAdd -h and -help for c_rehash script and app
Rich Salz [Mon, 12 Sep 2016 15:29:22 +0000 (11:29 -0400)]
Add -h and -help for c_rehash script and app

Resolves GH1515 and GH1509.

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 7d959c358a09244bcaea601121b276529ff437ce)

8 years agoAdd some sanity checks around usage of t_fromb64()
Matt Caswell [Thu, 8 Sep 2016 13:32:27 +0000 (14:32 +0100)]
Add some sanity checks around usage of t_fromb64()

The internal SRP function t_fromb64() converts from base64 to binary. It
does not validate that the size of the destination is sufficiently large -
that is up to the callers. In some places there was such a check, but not
in others.

Add an argument to t_fromb64() to provide the size of the destination
buffer and validate that we don't write too much data. Also add some sanity
checks to the callers where appropriate.

With thanks to Shi Lei for reporting this issue.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 73f0df8331910d6726d45ecaab12bd93cc48b4e2)

8 years agoAdd a test for 'openssl passwd'
Richard Levitte [Tue, 13 Sep 2016 21:23:51 +0000 (23:23 +0200)]
Add a test for 'openssl passwd'

Also, enlarge test group 20 to include openssl commands that aren't
tested otherwise

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 497f3bf9a75a2917e50b16b7985e87c89b86a39b)

8 years agoFix 'openssl passwd' with arguments -1 or -apr1
Richard Levitte [Tue, 13 Sep 2016 20:48:35 +0000 (22:48 +0200)]
Fix 'openssl passwd' with arguments -1 or -apr1

RT#4674

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 9f9f962d96425ed741569460791eee0280fcf942)

8 years agoRemove -xtarget=ultra from solaris(64)-sparcv9-cc builds.
Andy Grundman [Tue, 30 Aug 2016 21:25:10 +0000 (17:25 -0400)]
Remove -xtarget=ultra from solaris(64)-sparcv9-cc builds.

This flag got moved after -xarch=v9 in 1.1.0 and had the unexpected
side effect of the compiler building for 32-bit v8plusa instead of v9.

GH#1521
CLA: none; trivial

Signed-off-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit fa4618a2805e7115cf47d0cf0d15cb7b3c944bba)

8 years agobio.h: fix number of arguments passed to BIO_ptr_ctrl()
Viktor Szakats [Tue, 30 Aug 2016 16:56:49 +0000 (18:56 +0200)]
bio.h: fix number of arguments passed to BIO_ptr_ctrl()

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1520)
(cherry picked from commit c2efa78309e39019642f922b2babc1fd35f68768)

8 years agos_client: avoid warning on Windows/MS-DOS systems
Viktor Szakats [Tue, 30 Aug 2016 01:01:16 +0000 (03:01 +0200)]
s_client: avoid warning on Windows/MS-DOS systems

it appears when using gcc/mingw:
```
apps/s_client.c:815:9: warning: variable 'at_eof' set but not used [-Wunused-but-set-variable]
     int at_eof = 0;
         ^~~~~~
```
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1512)
(cherry picked from commit d485640b8083aea895ecf31b4ea6a338ee561b67)

8 years agoConfigure: detect gcc's dependency generation capability more accurately.
Andy Polyakov [Sat, 10 Sep 2016 19:12:56 +0000 (21:12 +0200)]
Configure: detect gcc's dependency generation capability more accurately.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 35c11bfc69e6b90fd1c4c4ca6ad3f500584ca939)

8 years agobn/bn_lcl.h: improve interoperability with clang and Android NDK.
Andy Polyakov [Sat, 10 Sep 2016 16:57:14 +0000 (18:57 +0200)]
bn/bn_lcl.h: improve interoperability with clang and Android NDK.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit cc2cb7bf63c62aaebd387f546a2fd673f367d9a8)

8 years agoutil/shlib_wrap.sh is now auto-generated so tell git to ignore it
Matt Caswell [Sun, 11 Sep 2016 16:47:39 +0000 (12:47 -0400)]
util/shlib_wrap.sh is now auto-generated so tell git to ignore it

Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 6462876f8d9f6538ad0fcd70717077d9ae174e56)

8 years agoAbort on unrecognised warning alerts
Matt Caswell [Mon, 12 Sep 2016 10:04:51 +0000 (11:04 +0100)]
Abort on unrecognised warning alerts

A peer continually sending unrecognised warning alerts could mean that we
make no progress on a connection. We should abort rather than continuing if
we receive an unrecognised warning alert.

Thanks to Shi Lei for reporting this issue.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 77a6be4dfc2ecf406c2559a99bea51317ce0f533)

8 years agoAdd enginesdir to libcrypto.pc pkg-config file
David Woodhouse [Sat, 27 Aug 2016 19:33:23 +0000 (20:33 +0100)]
Add enginesdir to libcrypto.pc pkg-config file

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit cdbbf9900253e8006868eba948248b1092a057de)

8 years agoFix EC_KEY_print so it prints out private key information
Tim Hudson [Fri, 9 Sep 2016 20:53:33 +0000 (06:53 +1000)]
Fix EC_KEY_print so it prints out private key information
even when the public key is not present in an EC_KEY

Reviewed-by: Stephen Henson <steve@openssl.org>
(cherry picked from commit 82f52631b21b021787f344733e08bb53f1e7f0ec)

8 years agoIf errno is ENXIO in BSS_new_file(), set BIO_R_NO_SUCH_FILE
Richard Levitte [Thu, 8 Sep 2016 21:39:26 +0000 (23:39 +0200)]
If errno is ENXIO in BSS_new_file(), set BIO_R_NO_SUCH_FILE

VMS sets that errno when the device part of a file spec is malformed
or a logical name that doesn't exist.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit e82e2186e93e9a678dd8c0c5ba084d21d27d4d62)

8 years agoTravis: Do asan and msan with shared libraries
Richard Levitte [Thu, 8 Sep 2016 08:57:31 +0000 (10:57 +0200)]
Travis: Do asan and msan with shared libraries

Reviewed-by: Emilia Käsper <emilia@openssl.org>
(cherry picked from commit 047a5da2494abf565f5fb76f5cd92a05cd7fadd4)

8 years agoFix mem leaks during auto-deinit
Matt Caswell [Mon, 5 Sep 2016 13:12:58 +0000 (14:12 +0100)]
Fix mem leaks during auto-deinit

Certain functions are automatically called during auto-deinit in order
to deallocate resources. However, if we have never entered a function which
marks lib crypto as inited then they never get called. This can happen if
the user only ever makes use of a small sub-set of functions that don't hit
the auto-init code.

This commit ensures all such resources deallocated by these functions also
init libcrypto when they are initially allocated.

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Ben Laurie <ben@openssl.org>
(cherry picked from commit 135648bcd0e9db029582d3d7627a90a1b566c5d6)

8 years agoEnsure trace recognises X25519
Matt Caswell [Fri, 2 Sep 2016 09:42:36 +0000 (10:42 +0100)]
Ensure trace recognises X25519

Using the -trace option to s_server or s_client was incorrectly printing
UNKNOWN for the X25519 curve.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 2d11f5b2ca863d4bd9e20b224932b247ed85842b)

8 years agoengines/afalg: make it compile with backward compatibility headers.
Andy Polyakov [Sun, 4 Sep 2016 20:53:22 +0000 (22:53 +0200)]
engines/afalg: make it compile with backward compatibility headers.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit 4e3997293b3825ed19de20d5484f8c0f66fce534)

8 years agoConfigurations/10-main.cf: AIX "facelift".
Andy Polyakov [Sat, 3 Sep 2016 20:03:55 +0000 (22:03 +0200)]
Configurations/10-main.cf: AIX "facelift".

Improve interchangeability of aix*-gcc targets by linking shared
libraries with -static-libgcc, and address linking problems with
vendor compiler.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit f780eaad5be140cf192191b8d79dc3671443e0b2)

8 years agochacha/asm/chacha-ppc.pl: add missing .text directive.
Andy Polyakov [Fri, 2 Sep 2016 22:46:01 +0000 (00:46 +0200)]
chacha/asm/chacha-ppc.pl: add missing .text directive.

RT#4667

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit abcbf7ed7ea249609d225e43e1bce3af1fa1028c)

8 years agoAdd missing debug strings.
Rich Salz [Wed, 7 Sep 2016 19:40:16 +0000 (15:40 -0400)]
Add missing debug strings.

Found by turning -Wswitch-enum on.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 252cfef151e302dce598503de09fd3a3341810d7)

8 years agoAllow asan, msan and ubsan to be configured with shared libraries
Richard Levitte [Wed, 7 Sep 2016 18:56:20 +0000 (20:56 +0200)]
Allow asan, msan and ubsan to be configured with shared libraries

The background story is that util/shlib_wrap.sh was setting LD_PRELOAD
or similar platform dependent variables, just in case the shared
libraries were built with -rpath.  Unfortunately, this doesn't work
too well with asan, msan or ubsan.

So, the solution is to forbid the combination of shared libraries,
-rpath and any of the sanity analyzers we can configure.

This changes util/shlib_wrap.sh so it only contains the code that sets
LD_PRELOAD when -rpath has been used when configuring.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 342a1a23793cb99921abeabe882adf8652ba715d)

8 years agoAvoid EVP_PKEY_cmp() crash on EC keys without public component
David Woodhouse [Wed, 7 Sep 2016 15:53:18 +0000 (16:53 +0100)]
Avoid EVP_PKEY_cmp() crash on EC keys without public component

Some hardware devices don't provide the public EC_POINT data. The only
way for X509_check_private_key() to validate that the key matches a
given certificate is to actually perform a sign operation and then
verify it using the public key in the certificate.

Maybe that can come later, as discussed in issue 1532. But for now let's
at least make it fail gracefully and not crash.

GH: 1532

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1547)

8 years agoConfigure: Reorganise the checking of disabled options
Richard Levitte [Tue, 6 Sep 2016 18:14:33 +0000 (20:14 +0200)]
Configure: Reorganise the checking of disabled options

The way we figured out what options are crypto algorithms and what are
something other was somewhat sketchy.  This change bases the
distinction on available sdirs instead.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 3e2dd30d665f3a312a45f945ffafb74ff6c420d6)

8 years agoGH1537: Avoid double-free in the EVP_PKEY API
Alex Gaynor [Wed, 7 Sep 2016 13:41:20 +0000 (09:41 -0400)]
GH1537: Avoid double-free in the EVP_PKEY API

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit d65c3615f6c658478503f4862f8055203a98038c)

8 years agoMisc BN fixes
Rich Salz [Mon, 5 Sep 2016 22:08:43 +0000 (18:08 -0400)]
Misc BN fixes

Never output -0; make "negative zero" an impossibility.
Do better checking on BN_rand top/bottom requirements and #bits.
Update doc.
Ignoring trailing garbage in BN_asc2bn.

Port this commit from boringSSL: https://boringssl.googlesource.com/boringssl/+/899b9b19a4cd3fe526aaf5047ab9234cdca19f7d%5E!/
        Ensure |BN_div| never gives negative zero in the no_branch code.

        Have |bn_correct_top| fix |bn->neg| if the input is zero so that we
        don't have negative zeros lying around.

        Thanks to Brian Smith for noticing.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 01c09f9fde5793e0b3712d602b02e2aed4908e8d)

8 years agoCorrect detection of group end in map file when testing symbol presence
Richard Levitte [Mon, 5 Sep 2016 22:48:13 +0000 (00:48 +0200)]
Correct detection of group end in map file when testing symbol presence

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 77a42b5f177e308233ab108806c48b9590a780e2)

8 years agoUnix build: have the makedepend and cc actions in one recipe
Richard Levitte [Sun, 4 Sep 2016 06:10:22 +0000 (08:10 +0200)]
Unix build: have the makedepend and cc actions in one recipe

In the case of using an independent makedepend, we had split that into
two separate recipes, one depending on the other.  However, there are
cases where the makedepend recipe was always trying, but doesn't
update the time stamp of the .d file because there are no actual
changes, and thereby causing constant updates of the object files.

This change makes one recipe that takes care of both makedepend och
cc, thereby avoiding these extra updates.

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit 7e5b8b93f2ffa8300b992d2e79c50f0e5266c61b)

8 years agomodes/asm/ghash-armv4.pl: improve interoperability with Android NDK.
Andy Polyakov [Sat, 27 Aug 2016 18:22:03 +0000 (20:22 +0200)]
modes/asm/ghash-armv4.pl: improve interoperability with Android NDK.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit 6cf412c473d8145562b76219ce3da73b201b3255)

8 years agoConfigurations/10-main.conf: fix omittions in commentary.
Andy Polyakov [Wed, 31 Aug 2016 18:50:08 +0000 (20:50 +0200)]
Configurations/10-main.conf: fix omittions in commentary.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit d5a39c12f0331dd0d14dbfcf2b98921be4cbb1da)

8 years agoConfigurations/10-main.conf: remove solaris-x86-cc target.
Andy Polyakov [Wed, 31 Aug 2016 14:13:10 +0000 (16:13 +0200)]
Configurations/10-main.conf: remove solaris-x86-cc target.

Since vendor assembler can't assemble our modules with -KPIC flag,
it, assembly support, was not available as an option. But this
means lack of side-channel resistant code, which is incompatible
with security by todays standards.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 216a0cc4d6a35a21e613f1e7e9eee957768bf9dd)

8 years agoMove 05-test_fuzz.t to 90-test_fuzz.t
Richard Levitte [Thu, 1 Sep 2016 19:05:34 +0000 (21:05 +0200)]
Move 05-test_fuzz.t to 90-test_fuzz.t

This adheres much better to the documentation in test/README

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 377ab6d183453dac5454abf62d5341a559caa264)

8 years agoRevert "INSTALL: add information on option no-fuzz-test"
Richard Levitte [Thu, 1 Sep 2016 11:44:05 +0000 (13:44 +0200)]
Revert "INSTALL: add information on option no-fuzz-test"

This reverts commit 7f9ae88817ddf0aac5c6bd95d9a5af1c54ed5bbf.

Reviewed-by: Emilia Käsper <emilia@openssl.org>
(cherry picked from commit 967e831e88bc45ea32e5b3c762945859996771ff)

8 years agoRevert "Make it possible to disable fuzz testing"
Richard Levitte [Thu, 1 Sep 2016 11:43:40 +0000 (13:43 +0200)]
Revert "Make it possible to disable fuzz testing"

This reverts commit eb40eaed727500bf4a15f848c99e37edd18e142e.

Reviewed-by: Emilia Käsper <emilia@openssl.org>
(cherry picked from commit a5e1f1230e09b249ff94cc48aeffd1b874cb937e)