For the same reasons as in the previous commit we must preserve errno
across dlopen calls. Some implementations (e.g. solaris) do not preserve
errno even on a successful dlopen call.

Fixes #6953
Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7680)

(cherry picked from commit 3cb4e7dc1cf92022f62b9bbdd59695885a1265ff)

This function can end up being called during ERR_get_error() if we are
initialising. ERR_get_error() must preserve errno since it gets called via
SSL_get_error(). If that function returns SSL_ERROR_SYSCALL then you are
supposed to inspect errno.
Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7680)

(cherry picked from commit 71b1ceffc4c795f5db21861dd1016fbe23a53a53)

Reviewed-by: NKurt Roeckx <kurt@roeckx.be>
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7852)

(cherry picked from commit f2f734d4f9e34643a1d3e5b79d2447cd643519f8)

Reviewed-by: NKurt Roeckx <kurt@roeckx.be>
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7852)

(cherry picked from commit a8bf2f8f24bda57c860e4356f31264fff63fe8b9)

Reviewed-by: NKurt Roeckx <kurt@roeckx.be>
Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7850)

(cherry picked from commit 91d0fd1c2753f0f7d6e0953eed3cfb6eb96d8ff4)

It turns out that the strictness that was implemented in
EVP_PKEY_asn1_new() (see Github openssl/openssl#6880) was badly placed
for some usages, and that it's better to do this check only when the
method is getting registered.

Fixes #7758
Reviewed-by: NTim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7847)

(cherry picked from commit a86003162138031137727147c9b642d99db434b1)

Reviewed-by: NPaul Yang <yang.yang@baishancloud.com>
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7705)

(cherry picked from commit 56bc3ea0a49b8aa99a4b031e7cb48ea87fe28e2d)

Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7750)

(cherry picked from commit 425dde5d75a6ffff9a1027107711b99da39e3519)

Fixes #7698
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7750)

(cherry picked from commit 3a17b9a46ecf54a30073534dff04e3223ff98813)

Fix some issues in tls13_hkdf_expand() which impact the above function
for TLSv1.3. In particular test that we can use the maximum label length
in TLSv1.3.
Reviewed-by: NTim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7755)

(cherry picked from commit 0fb2815b873304d145ed00283454fc9f3bd35e6b)

This reverts commit ec0c5f5693e39c5a013f81e6dd9dfd09ec65162d.

SSL_export_keying_material() may use longer label lengths.

Fixes #7712
Reviewed-by: NTim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7755)

(cherry picked from commit ed371b8cbac0d0349667558c061c1ae380cf75eb)

Fixed-top interfaces tolerate zero-padded inputs and facilitate
constant-time-ness. bn_div_fixed_top tolerates zero-padded dividend,
but not divisor. It's argued that divisor's length is public even
when value is secret.

[extended tests]
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7589)

(cherry picked from commit 3a4a88f436ed1dd1165e0b59c1ca4a25e9e1d690)

and add template for constant-time bn_div_3_words.
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7589)

(cherry picked from commit 3da2e9c4ee45989a426ff513dc6c6250d1e460de)

Reviewed-by: NPaul Dale <paul.dale@oracle.com>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7589)

(cherry picked from commit 31703da3b8e743c07a9fdb6025aa526594c2a319)

It's being replaced with constant-time alternative.
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7589)

(cherry picked from commit b34446cca2b1814fa674adc8076707d2b65a697f)

Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7697)

(cherry picked from commit e65a79d2aae76c27d68587ff66cb6b65a3283e50)

Check that s is less than the order before attempting to verify the
signature as per RFC8032 5.1.7

Fixes #7693
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7697)

(cherry picked from commit 0ac8f35c04c4fcdee421170ae6351e42b9e84ce4)

Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7744)

(cherry picked from commit 7b4a3515a4ddb567d48000e61d7cb640d0c5f261)

Copy of RSA_padding_check_PKCS1_type_2 with a twist that rejects padding
if nul delimiter is preceded by 8 consecutive 0x03 bytes.
Reviewed-by: NRichard Levitte <levitte@openssl.org>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(cherry picked from commit 603221407ddc6404f8c417c6beadebf84449074c)

And make RSAErr call unconditional.
Reviewed-by: NRichard Levitte <levitte@openssl.org>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(cherry picked from commit 75f5e944be97f28867e7c489823c889d89d0bd06)

And make RSAErr call unconditional.
Reviewed-by: NRichard Levitte <levitte@openssl.org>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(cherry picked from commit e875b0cf2f10bf2adf73e0c2ec81428290f4660c)

Reviewed-by: NRichard Levitte <levitte@openssl.org>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(cherry picked from commit 89072e0c2a483f2ad678e723e112712567b0ceb1)

Expected usage pattern is to unconditionally set error and then
wipe it if there was no actual error.
Reviewed-by: NRichard Levitte <levitte@openssl.org>
Reviewed-by: NMatt Caswell <matt@openssl.org>
(cherry picked from commit f658a3b64d8750642f4975090740865f770c2a1b)

Config options 'no-err' and 'no-autoerrinit'
Reviewed-by: NMatt Caswell <matt@openssl.org>
Reviewed-by: NTim Hudson <tjh@openssl.org>
Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7711)

(cherry picked from commit f1d49ed94707f5f0799541aa8e528dce67712948)

If compile OpenSSL with SSL_DEBUG macro, some test cases will cause the
process crashed in the debug code.
Reviewed-by: NTim Hudson <tjh@openssl.org>
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7707)

(cherry picked from commit 5a4481f0e0bbf836c7f41805e84617784cb6b1db)

Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7703)

(cherry picked from commit f19d20b60fcdfde802f96afa3796f20d0da25eab)

We only convert lowercase .s to .asm, that turned out not to be sufficient.
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7703)

(cherry picked from commit c739e676ebc6e7f66aabee1ac3d2040b3103a890)

Updated "condition" logic lifted from Theo Buehler's LibreSSL commit https://github.com/libressl-portable/openbsd/commit/517358603b4be76d48a50007a0d414c2072697ddReviewed-by: NMatt Caswell <matt@openssl.org>
Reviewed-by: NNicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/7619)

(cherry picked from commit 900fd8f375ca758d182e894bc1556509b231dbc8)

There's too high a chance that the openssl app and perl get different
messages for some error numbers.

[extended tests]
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7699)

(cherry picked from commit 0777de15ff09105593c7e778cf47f97a9144d273)

Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7696)

(cherry picked from commit 3be389435fc7b94623d972b622dbd9f0cd5c34f7)

Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7696)

(cherry picked from commit b741f153b2f24139d7210b1b0c9caf561f4900e8)

Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7696)

(cherry picked from commit 76bc401cc63219a462224884cb4af787e17725ed)

When creating a tarball, it's pointless to include scripts that assume
a git workspace.
Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7696)

(cherry picked from commit b9a694717902af796639e1dff641ba620703303b)

Also adds missing copyright boilerplate to util/mktar.sh
Reviewed-by: NMatthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7696)

(cherry picked from commit b42922ea2f605fd6c42faad1743fb27be5f7f1f3)

This typo prevented ia64 assembler to be compiled on VMS
Reviewed-by: NTim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7700)

(cherry picked from commit 04375857cf4971460c473fa44bef86351f2177fe)

It turns out that on VMS, strerror() returns messages with added
spaces at the end.

We wouldn't had noticed if it wasn't for perl trimming those spaces
off for its own sake and thereby having test/recipes/02-test_errstr.t
fail on VMS.

The safe fix is to do the same trimming ourselves.
Reviewed-by: NPaul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7701)

(cherry picked from commit 9f15e5b911ba6053e09578f190354568e01c07d7)

If the private key says it can only support one specific digest, then
don't ask it to perform a different one.

Fixes: #7348
Reviewed-by: NNicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: NRichard Levitte <levitte@openssl.org>

(cherry picked from commit 2d263a4a73f852005b16359873475d48755999ad)
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7609)

ASN1_PKEY_CTRL_DEFAULT_MD_NID is documented to return 2 for a mandatory
digest algorithm, when the key can't support any others. That isn't true
here, so return 1 instead.

Partially fixes #7348
Reviewed-by: NNicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: NRichard Levitte <levitte@openssl.org>

(cherry picked from commit eb7eb1378cd15c4652884b3701d4c0ef27b5b8a6)
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7609)

Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7692)

(cherry picked from commit 8d9535ec3e317641b8e551973c8cfe2ee1c89296)

Since recently, OpenSSL tarballs are produced with 'make tar' rather
than 'make dist', as the latter has turned out to be more troublesome
than useful.

The next step to look at is why we would need to configure at all to
produce a Makefile just to produce a tarball.  After all, the tarball
should now only contain source files that are present even without
configuring.

Furthermore, the current method for producing tarballs is a bit
complex, and can be greatly simplified with the right tools.  Since we
have everything versioned with git, we might as well use the tool that
comes with it.

Added: util/mktar.sh, a simple script to produce OpenSSL tarballs.  It
takes the options --name to modify the prefix of the distribution, and
--tarfile tp modify the tarball file name specifically.

This also adds a few entries in .gitattributes to specify files that
should never end up in a distribution tarball.
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7692)

(cherry picked from commit 8c209eeef426ded66ce99048f535f35d08b88462)