]> code.ossystems Code Review - openembedded-core.git/log
openembedded-core.git
7 years agobusybox: ifupdown:pass interface device name for ipv6 route command
Haiqing Bai [Wed, 8 Mar 2017 09:27:30 +0000 (17:27 +0800)]
busybox: ifupdown:pass interface device name for ipv6 route command

IPv6 routes need the device argument for link-local routes, or they
cannot be used at all. E.g. "gateway fe80::def" seems to be used in
some places, but kernel refuses to insert the route unless device
name is explicitly specified in the route addition.

Signed-off-by: Haiqing Bai <Haiqing.Bai@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 96ed437d57316153453bb5e170a4fd4f3a95883d)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agobusybox: allow libiproute to handle table ids larger than 255
Lukasz Nowak [Tue, 13 Dec 2016 16:19:28 +0000 (16:19 +0000)]
busybox: allow libiproute to handle table ids larger than 255

These changes are required for compatibility with ConnMan, which by default
uses table ids greater than 255.

Signed-off-by: Lukasz Nowak <lnowak@tycoint.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit e9114bdd8a83b88f59526780910c49e3092fdd57)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agobinutils: Fix CVE-2017-6965 and CVE-2017-6966
Yuanjie Huang [Tue, 11 Apr 2017 01:58:34 +0000 (18:58 -0700)]
binutils: Fix CVE-2017-6965 and CVE-2017-6966

Backport upstream commit to address vulnerabilities:

CVE: CVE-2017-6965
[BZ 21137] -- https://sourceware.org/bugzilla/show_bug.cgi?id=21137

Fix readelf writing to illegal addresses whilst processing corrupt input
files containing symbol-difference relocations.

PR binutils/21137
* readelf.c (target_specific_reloc_handling): Add end parameter.
Check for buffer overflow before writing relocated values.
(apply_relocations): Pass end to target_specific_reloc_handling.

CVE: CVE-2017-6966
[BZ 21139] -- https://sourceware.org/bugzilla/show_bug.cgi?id=21139

Fix read-after-free error in readelf when processing multiple, relocated
sections in an MSP430 binary.

PR binutils/21139
* readelf.c (target_specific_reloc_handling): Add num_syms
parameter.  Check for symbol table overflow before accessing
symbol value.  If reloc pointer is NULL, discard all saved state.
(apply_relocations): Pass num_syms to target_specific_reloc_handling.
Call target_specific_reloc_handling with a NULL reloc pointer
after processing all of the relocs.

Signed-off-by: Yuanjie Huang <yuanjie.huang@windriver.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agobase-files: resize only serial tty's in profile
Daniel Díaz [Fri, 7 Apr 2017 21:33:45 +0000 (16:33 -0500)]
base-files: resize only serial tty's in profile

[Backported from master.]

We don't want to run resize on non serial consoles. There's
been an earlier attempt (6557787), so this builds upon that.

The problem we're seeing is that if there is text buffered in
the virtual console (like from a desperate user trying to
enter login details), resize will get stuck while calling
  ioctl(tty, TCSETAW);

Since serial consoles are named (not just numbered), this
change limits resize's reach even further to run only on
/dev/tty[A-z] (thus avoiding /dev/tty[0-9]).

Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agologrotate: replace fedorahosted.org SRC_URI with github.com source
Choong YinThong [Tue, 18 Apr 2017 19:57:05 +0000 (19:57 +0000)]
logrotate: replace fedorahosted.org SRC_URI with github.com source

fedorahosted.org was retired on March 1st, 2017. This is to
update the SRC_URI to point to github.com.
Update the ${PN} to ${BPN} in order to pass the autobuilder
mulitlib enable configuration.

[YOCTO #11226]

Signed-off-by: Choong YinThong <yin.thong.choong@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agovolatile-binds: correct some errors reported by systemd
Joe Slater [Thu, 30 Mar 2017 21:16:08 +0000 (14:16 -0700)]
volatile-binds: correct some errors reported by systemd

systemd-tmpfiles-setup will fail at boot, so we suppress
the default versions of etc.conf and home.conf.

We also make sure that /var/{cache,spool} and /srv are writeable
if they exist.

Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agoparselogs: Whitelist NUC6 firmware load error message on genericx86-64
Sullivan, California L [Mon, 27 Mar 2017 21:33:37 +0000 (14:33 -0700)]
parselogs: Whitelist NUC6 firmware load error message on genericx86-64

This was already whitelisted, but the 4.8 kernel changed the error
message, causing it to get caught by parselogs again.

Fixes [YOCTO #10494].

Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit e1bad14231115f3b1a2bf844ef5b2022c648b55d)
Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agosstate.bbclass: update .siginfo atime
Ed Bartosh [Mon, 20 Mar 2017 17:38:20 +0000 (13:38 -0400)]
sstate.bbclass: update .siginfo atime

.siginfo files are not being accessed from local or NFS-mounted
sstate mirrors when sstate package is installed, so their atime
is not updated. If sstate mirror is cleaned based on access time,
they get deleted, even though they are still being used.

Updated atime of .siginfo symlinks with 'touch -a'. This command
dereferences symlinks pointing to the local mirror and updates
atime of the .siginfo file on the mirror.

[YOCTO #10857]

Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agorootfs.py: Respect OPKGLIBDIR variable
Martin Jansa [Thu, 23 Mar 2017 12:16:29 +0000 (13:16 +0100)]
rootfs.py: Respect OPKGLIBDIR variable

* when OPKGLIBDIR doesn't have the default /var/lib value it will
  silently fail to copy package database from normal rootfs to debugfs
  rootfs and then when trying to install *-dbg complimentary packages
  it won't install anything, because installed_pkgs.txt file generated
  from debugfs is empty

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agopackage_manager.py: respect OPKGLIBDIR
Martin Jansa [Thu, 23 Mar 2017 12:16:30 +0000 (13:16 +0100)]
package_manager.py: respect OPKGLIBDIR

* respect it for incremental rootfs generation
* add lists_dir option to opkg.conf
* also fix setting info_dir and status_file when they use default value, the
  problem is that self.opkg_dir is already prefixed with rootfs directory,
  comparing it with /var/lib/opkg always returned false and the options were
  appended to config file unnecessary
* with opkg 0.3.4 we can use VARDIR prefix added in:
  commit d2a8e23dc669adc398f4bb8bcfcabfcf925708f7
  Author: Florin Gherendi <floring2502@gmail.com>
  Date:   Mon Dec 19 12:25:38 2016 +0200
  libopkg: make the /var and /etc directories configurable at compile time.

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agopython-3-manifest: fix adding imp to importlib
Anders Darander [Mon, 13 Feb 2017 22:06:56 +0000 (16:06 -0600)]
python-3-manifest: fix adding imp to importlib

Commit: 512334f102a33833d39af53467894315f0715d07
"python-3.5-manifest: Add imp to importlib"

added imp to importlib in the generated manifest, but not in the generator script.

Signed-off-by: Anders Darander <anders@chargestorm.se>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agorunqemu: support multiple qemus running when nfs
Robert Yang [Wed, 23 Nov 2016 08:57:39 +0000 (00:57 -0800)]
runqemu: support multiple qemus running when nfs

Fixed:
* In build1:
  $ runqemu nfs qemux86-64
  In build2:
  $ runqemu nfs qemux86-64

  It would fail before since the port numerbs and conf files are
  conflicted, now make runqemu-export-rootfs work together with runqemu to
  fix the problem.

* And we don't need export PSEUDO_LOCALSTATEDIR in runqemu, the
  runqemu-export-rootfs can handle it well based on NFS_EXPORT_DIR.

* Remove "async" option from unfsd to fix warning in syslog:
  Warning: unknown exports option `async' ignored

* Fixed typos

Both slirp and tap can work.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agorunqemu-export-rootfs: fix inconsistent var names
Robert Yang [Wed, 23 Nov 2016 08:00:18 +0000 (00:00 -0800)]
runqemu-export-rootfs: fix inconsistent var names

Fixed:
$ runqemu nfs qemux86-64
[snip]
On your target please remember to add the following options for NFS
nfsroot=IP_ADDRESS:/path/to/nfsroot,nfsvers=3,port=,mountprog=,nfsprog=,udp,mountport=
[snip]

Note that the values are null, this is because their var names are
inconsistent.

[YOCTO #10519]

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agosanity.bbclass: modify check for shell
Juro Bystricky [Wed, 15 Mar 2017 19:39:12 +0000 (12:39 -0700)]
sanity.bbclass: modify check for shell

Due to the recently implemented update-alternatives for bash binary,
sanity checker may end up with a (false-positive) error such as:

Error, /bin/sh links to /bin/bash.bash, must be dash or bash

This patch modifies the test: presence of "/bash" or "/dash" in shell binary
name results in pass.

[YOCTO#11108]

Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agoimage_types.bbclass: fix image dependency chain collection
Andre McCurdy [Tue, 21 Feb 2017 18:04:50 +0000 (10:04 -0800)]
image_types.bbclass: fix image dependency chain collection

If image type "foo" depends on image type "bar.xz", then dependencies
should be collected from the base image type (ie "IMAGE_DEPENDS_bar")
not from "IMAGE_DEPENDS_bar.xz".

Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 8a9f249a9166347cc0468191ce130003e3d306e1)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agoimage_types.bbclass: IMAGE_TYPEDEP_ now adds deps for conversion types
Randy Witt [Tue, 21 Feb 2017 18:04:49 +0000 (10:04 -0800)]
image_types.bbclass: IMAGE_TYPEDEP_ now adds deps for conversion types

Previously if IMAGE_TYPEDEP_* contained a conversion type of the form,
"foo.bar", the dependency on CONVERSION_DEPENDS_bar would not get added
to the task depends for do_rootfs.

[YOCTO #10883]

Signed-off-by: Randy Witt <randy.e.witt@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 037d39898e0e16c6d5b24a8d3844abfb328d3c14)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agogmp: Disable assembly for MIPS R6
Zubair Lutfullah Kakakhel [Wed, 23 Nov 2016 13:47:31 +0000 (13:47 +0000)]
gmp: Disable assembly for MIPS R6

gmplib has some assembly routines. These have not been optimized
for MIPS R6 yet. Add --disable-assembly so that the C implementation
is used.

Signed-off-by: Zubair Lutfullah Kakakhel <Zubair.Kakakhel@imgtec.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit b0975809fe43b7506cc30a60245f5e32b275a3ec)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agoopenssl: Add support for many MIPS configurations
Zubair Lutfullah Kakakhel [Fri, 16 Dec 2016 15:14:19 +0000 (15:14 +0000)]
openssl: Add support for many MIPS configurations

Add more case statements to catch MIPS tune configurations

Signed-off-by: Zubair Lutfullah Kakakhel <Zubair.Kakakhel@imgtec.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit cd1f6fbf9a2113cf510c25de2eb3895468e79149)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agobusybox: refresh the flock patch
Maxin John [Tue, 17 Jan 2017 16:20:34 +0000 (18:20 +0200)]
busybox: refresh the flock patch

Upstream accepted the flock fix with some improvements. Backport those
changes.

Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Maxin B. John <maxin.john@intel.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agowic: plugins: rawcopy: Fixed wrong variable type
Daniel Schultz [Tue, 7 Mar 2017 13:41:55 +0000 (14:41 +0100)]
wic: plugins: rawcopy: Fixed wrong variable type

Without the int() function this variable will be a string. This will led
to a error in Filemap on line 545 due wrong types.

> [...]
>   File
> ".../poky/scripts/lib/wic/filemap.py", line 545, in sparse_copy
>     if start < skip < end:
> TypeError: unorderable types: int() < str()

Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 46b5814bcdc0e7e3cb293e877e2aa949baf5fef8)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agowic: filemap: Fixed spared_copy skip
Daniel Schultz [Tue, 7 Mar 2017 13:41:56 +0000 (14:41 +0100)]
wic: filemap: Fixed spared_copy skip

This patches removes the empty space in front of the copied file which
was skipped. Without this reduction it's not possible to place a
partition with rawcopy and skip parameter on a desired alignment.

Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 5c024d71f9413b81ee1707dbc41f0721f8f27bdb)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agotzcode-native: Set cc to ${CC}
Richard Purdie [Thu, 9 Mar 2017 09:48:44 +0000 (09:48 +0000)]
tzcode-native: Set cc to ${CC}

Building on a system without "cc" showed this recipe doesn't respect
the $CC variable. Fix this by passing the right option to the makefile.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 402080c0a77443f541fa3d658b79f3fba327279d)

7 years agotzdata: update to 2017a
Armin Kuster [Wed, 1 Mar 2017 03:48:31 +0000 (19:48 -0800)]
tzdata: update to 2017a

 Briefly: Southern Chile moves from -04/-03 to -03, and Mongolia
  discontinues DST.

  Changes to future time stamps

    Mongolia no longer observes DST.  (Thanks to Ganbold Tsagaankhuu.)

    Chile's Region of Magallanes moves from -04/-03 to -03 year-round.
    Its clocks diverge from America/Santiago starting 2017-05-13 at
    23:00, hiving off a new zone America/Punta_Arenas.  Although the
    Chilean government says this change expires in May 2019, for now
    assume it's permanent.  (Thanks to Juan Correa and Deborah
    Goldsmith.)  This also affects Antarctica/Palmer.

  Changes to past time stamps

    Fix many entries for historical time stamps for Europe/Madrid
    before 1979, to agree with tables compiled by Pere Planesas of the
    National Astronomical Observatory of Spain.  As a side effect,
    this changes some time stamps for Africa/Ceuta before 1929, which
    are probably guesswork anyway.  (Thanks to Steve Allen and
    Pierpaolo Bernardi for the heads-ups, and to Michael Deckers for
    correcting the 1901 transition.)

    Ecuador observed DST from 1992-11-28 to 1993-02-05.
    (Thanks to Alois Treindl.)

    Asia/Atyrau and Asia/Oral were at +03 (not +04) before 1930-06-21.
    (Thanks to Stepan Golosunov.)

  Changes to past and future time zone abbreviations

    Switch to numeric time zone abbreviations for South America, as
    part of the ongoing project of removing invented abbreviations.
    This avoids the need to invent an abbreviation for the new Chilean
    new zone.  Similarly, switch from invented to numeric time zone
    abbreviations for Afghanistan, American Samoa, the Azores,
    Bangladesh, Bhutan, the British Indian Ocean Territory, Brunei,
    Cape Verde, Chatham Is, Christmas I, Cocos (Keeling) Is, Cook Is,
    Dubai, East Timor, Eucla, Fiji, French Polynesia, Greenland,
    Indochina, Iran, Iraq, Kiribati, Lord Howe, Macquarie, Malaysia,
    the Maldives, Marshall Is, Mauritius, Micronesia, Mongolia,
    Myanmar, Nauru, Nepal, New Caledonia, Niue, Norfolk I, Palau,
    Papua New Guinea, the Philippines, Pitcairn, Qatar, Réunion, St
    Pierre & Miquelon, Samoa, Saudi Arabia, Seychelles, Singapore,
    Solomon Is, Tokelau, Tuvalu, Wake, Vanuatu, Wallis & Futuna, and
    Xinjiang; for 20-minute daylight saving time in Ghana before 1943;
    for half-hour daylight saving time in Belize before 1944 and in
    the Dominican Republic before 1975; and for Canary Islands before
    1946, for Guinea-Bissau before 1975, for Iceland before 1969, for
    Indian Summer Time before 1942, for Indonesia before around 1964,
    for Kenya before 1960, for Liberia before 1973, for Madeira before
    1967, for Namibia before 1943, for the Netherlands in 1937-9, for
    Pakistan before 1971, for Western Sahara before 1977, and for
    Zaporozhye in 1880-1924.

    For Alaska time from 1900 through 1967, instead of "CAT" use the
    abbreviation "AST", the abbreviation commonly used at the time
    (Atlantic Standard Time had not been standardized yet).  Use "AWT"
    and "APT" instead of the invented abbreviations "CAWT" and "CAPT".

    Use "CST" and "CDT" instead of invented abbreviations for Macau
    before 1999 and Taiwan before 1938, and use "JST" instead of the
    invented abbreviation "JCST" for Japan and Korea before 1938.

  Change to database entry category

    Move the Pacific/Johnston link from 'australasia' to 'backward',
    since Johnston is now uninhabited.

Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 784a8379260759700ab8482b1699c9567d6e9f3f)

7 years agotzcode-native: update to 2017a
Armin Kuster [Wed, 1 Mar 2017 03:48:30 +0000 (19:48 -0800)]
tzcode-native: update to 2017a

  Changes to code

    zic no longer mishandles some transitions in January 2038 when it
    attempts to work around Qt bug 53071.  This fixes a bug affecting
    Pacific/Tongatapu that was introduced in zic 2016e. localtime.c
    now contains a workaround, useful when loading a file generated by
    a buggy zic.  (Problem and localtime.c fix reported by Bradley
    White.)

    zdump -i now outputs non-hour numeric time zone abbreviations
    without a colon, e.g., "+0530" rather than "+05:30".  This agrees
    with zic %z and with common practice, and simplifies auditing of
    zdump output.

    zdump is now buildable again with -DUSE_LTZ=0.
    (Problem reported by Joseph Myers.)

    zdump.c now always includes private.h, to avoid code duplication
    with private.h.  (Problem reported by Kees Dekker.)

    localtime.c no longer mishandles early or late timestamps
    when TZ is set to a POSIX-style string that specifies DST.
    (Problem reported by Kees Dekker.)

    date and strftime now cause %z to generate "-0000" instead of
    "+0000" when the UT offset is zero and the time zone abbreviation
    begins with "-".

  Changes to documentation and commentary

    The 'Theory' file now better documents choice of historical time
    zone abbreviations.  (Problems reported by Michael Deckers.)

    tz-link.htm now covers leap smearing, which is popular in clouds.

Signed-off-by: Armin Kuster <akuster@mvista.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 6c95fbf51ec538e29083a4a890d106b732c1b182)

7 years agotzcode-native: update to 2016j
Armin Kuster [Thu, 15 Dec 2016 15:21:44 +0000 (07:21 -0800)]
tzcode-native: update to 2016j

Changes to code

    zic no longer mishandles file systems that lack hard links, fixing
    bugs introduced in 2016g.  (Problems reported by Tom Lane.)
    Also, when the destination already contains symbolic links, zic
    should now work better on systems where the 'link' system call
    does not follow symbolic links.

Changes to documentation and commentary

    tz-link.htm now documents the relationship between release version
    numbers and development-repository commit tags.  (Suggested by
    Paul Koning.)

    The 'Theory' file now documents UT.

    iso3166.tab now accents "Curaçao", and commentary now mentions
the names "Cabo Verde" and "Czechia". (Thanks to Jiří Boháč.)

(From OE-Core rev: 3fb5ddce97af1d4ada0dcc7f9c0ceef6ac392918)

Signed-off-by: Armin Kuster <akuster808@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit f9d6b482f2275c054dff91b81ac901b612f9c9b4)

7 years agotzdata: update to 2016j
Armin Kuster [Thu, 15 Dec 2016 15:21:45 +0000 (07:21 -0800)]
tzdata: update to 2016j

Briefly: Saratov, Russia moves from +03 to +04 on 2016-12-04.

  Changes to future time stamps

    Saratov, Russia switches from +03 to +04 on 2016-12-04 at 02:00.
    This hives off a new zone Europe/Saratov from Europe/Volgograd.
    (Thanks to Yuri Konotopov and Stepan Golosunov.)

  Changes to past time stamps

    The new zone Asia/Atyrau for Atyraū Region, Kazakhstan, is like
    Asia/Aqtau except it switched from +05/+06 to +04/+05 in spring
    1999, not fall 1994.  (Thanks to Stepan Golosunov.)

  Changes to past time zone abbreviations

    Asia/Gaza and Asia/Hebron now use "EEST", not "EET", to denote
summer time before 1948. The old use of "EET" was a typo.

(From OE-Core rev: a0b389b5f3effad5194109f9914dec6125b3569d)

Signed-off-by: Armin Kuster <akuster808@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 147d7802926153ef0abb12d5e738576d8dfd02d7)

7 years agotar: CVE-2016-6321
Sona Sarmadi [Fri, 24 Feb 2017 12:49:18 +0000 (13:49 +0100)]
tar: CVE-2016-6321

Skip members whose names contain "..".

Reference:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-6321

Upstream patch:
http://git.savannah.gnu.org/cgit/tar.git/commit/?id=7340f67b9860ea0531c1450e5aa261c50f671

Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit cfa2b5facd1aa6a2bac4cb04687e1a977c533934)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
7 years agobash: fix CVE-2016-9401
Li Zhou [Mon, 13 Feb 2017 02:53:24 +0000 (10:53 +0800)]
bash: fix CVE-2016-9401

popd in bash might allow local users to bypass the restricted shell
and cause a use-after-free via a crafted address.

Porting patch from <https://ftp.gnu.org/pub/gnu/bash/bash-4.4-patches/
bash44-006> to solve CVE-2016-9401.

Signed-off-by: Li Zhou <li.zhou@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 6987b317d5ce8dc50a37ebba395aa8424bec358c)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agoRevert "file: update SRCREV for 5.28 to fix fetch fail on missing commit"
Richard Purdie [Tue, 21 Mar 2017 22:17:57 +0000 (22:17 +0000)]
Revert "file: update SRCREV for 5.28 to fix fetch fail on missing commit"

This reverts commit 87c17b352daa63bc4f62e8802982288ab1839f1b.

Upstream restored the original hashes.

8 years agofile: update SRCREV for 5.28 to fix fetch fail on missing commit
Paul Gortmaker [Fri, 17 Mar 2017 23:24:00 +0000 (19:24 -0400)]
file: update SRCREV for 5.28 to fix fetch fail on missing commit

Machines that cloned a while ago will have the commit, but new
deployments won't because it seems the upstream changed/rebased
and the old commit ID has been garbage-collected away.  Hence
the fetch fails to check out the named commit ID.

Both the old (gone) commit, and the "new" commit show the same
dates and commit log and point at 5.28, so hopefully this is
the right thing to do.  A git diff of the two seems to only show
a blanket uprev of CVS tags and deletion of a couple autogen'd
files, and no real source changes.

(From OE-Core rev: adb71e06768adadda7b69c3b5e81ca3ad67237f4)

Cc: Christos Zoulas <christos@zoulas.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/oe/gpg_sign: fix rpm signing with gpg > 2.1
Markus Lehtonen [Fri, 3 Mar 2017 12:37:27 +0000 (14:37 +0200)]
lib/oe/gpg_sign: fix rpm signing with gpg > 2.1

We need to check the gpg version and alter its command line options
accordingly.

[YOCTO #11054]

Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 44a44b7e582a5a654baf21829d168568481c13d9)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/oe/gpg_sign: make gpg version a property of the signer
Markus Lehtonen [Fri, 3 Mar 2017 12:37:26 +0000 (14:37 +0200)]
lib/oe/gpg_sign: make gpg version a property of the signer

Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit a00a362e3dc18ba04230cbbd6f91264e5d76f40d)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agorpm: support customizing gpg command line
Markus Lehtonen [Fri, 3 Mar 2017 12:37:25 +0000 (14:37 +0200)]
rpm: support customizing gpg command line

Add a new %_gpg_sign_cmd_extra_args macro that allows customizing the
gpg options used when signing rpm packages. This is needed to be able to
sign packages with gpg 2.1 which requires "--pinentry-mode loopback" to
allow non-interactive signing.

[YOCTO #11054]

Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 373a7146d596d27376a003014df0d06f3df5348d)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/oe/gpg_sign: sign rpm packages in chunks of 100
Markus Lehtonen [Fri, 3 Mar 2017 12:37:24 +0000 (14:37 +0200)]
lib/oe/gpg_sign: sign rpm packages in chunks of 100

Split the file list into chunks in order to avoid
"OSError: [Errno 7] Argument list too long"

This would happend when a package has huge amount of subpackages, e.g.
glibc-locale.

[YOCTO #11069]

Signed-off-by: Markus Lehtonen <markus.lehtonen@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit 874f5016fd4dc76bc867b68470297fe59e78a9e6)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoe2fsprogs: have configure expand @mkdir_p@
Joe Slater [Tue, 7 Mar 2017 23:00:55 +0000 (15:00 -0800)]
e2fsprogs: have configure expand @mkdir_p@

Without doing this, locale data is not written
into usr/src/locale.

We could also just use MKDIR_P instead of mkdir_p.

Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agopython-3.5-manifest: Add imp to importlib
Juro Bystricky [Mon, 13 Mar 2017 23:06:17 +0000 (16:06 -0700)]
python-3.5-manifest: Add imp to importlib

The imp python module is the forerunner of importlib. Include imp in
the importlib subpackage instead of the misc subpackage so that it can
be depended on without bringing in a bunch of unrelated, unused modules.

(Slightly modified backport of afd9de380dc04780da1eeb94a4c8c46529254231
by George McCollister <george.mccollister@gmail.com>)

[YOCTO#11134]

Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agogdb-cross-canadian: Depend on nativesdk-python3-importlib
George McCollister [Mon, 13 Mar 2017 23:06:16 +0000 (16:06 -0700)]
gdb-cross-canadian: Depend on nativesdk-python3-importlib

Add missing dependency on nativesdk-python3-importlib so the imp Python
module is installed.

Before this patch, running gdb from the sdk would give the following
error:

Python Exception <class 'ImportError'> No module named 'imp':

(From OE-Core rev: ad00a31fbe93e073a2d83616efcd08c5a7ef37c9)

Signed-off-by: George McCollister <george.mccollister@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobuild-appliance-image: Update to morty head revision
Richard Purdie [Wed, 8 Feb 2017 12:00:29 +0000 (12:00 +0000)]
build-appliance-image: Update to morty head revision

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoopenssl: CVE: CVE-2017-3731
Alexandru Moise [Tue, 7 Feb 2017 11:48:47 +0000 (13:48 +0200)]
openssl: CVE: CVE-2017-3731

If an SSL/TLS server or client is running on a 32-bit host, and a
specific cipher is being used, then a truncated packet can cause that
server or client  to perform an out-of-bounds read, usually resulting
in a crash.

Backported from:
https://github.com/openssl/openssl/commit/8e20499629b6bcf868d0072c7011e590b5c2294d
https://github.com/openssl/openssl/commit/2198b3a55de681e1f3c23edb0586afe13f438051

* CVE: CVE-2017-3731

Upstream-status: Backport

Signed-off-by: Alexandru Moise <alexandru.moise@windriver.com>
8 years agolibx11: CVE-2016-7943
Sona Sarmadi [Mon, 30 Jan 2017 11:46:23 +0000 (12:46 +0100)]
libx11: CVE-2016-7943

The XListFonts function in X.org libX11 before 1.6.4 might allow
remote X servers to gain privileges via vectors involving length
fields, which trigger out-of-bounds write operations.

References:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-7943
https://lists.x.org/archives/xorg-announce/2016-October/002720.html

Upstream patch:
https://cgit.freedesktop.org/xorg/lib/libX11/commit/?id=8c29f1607a31dac0911e45a0dd3d74173822b3c9

Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
8 years agolibX11: CVE-2016-7942
Sona Sarmadi [Mon, 30 Jan 2017 11:46:22 +0000 (12:46 +0100)]
libX11: CVE-2016-7942

The XGetImage function in X.org libX11 before 1.6.4 might allow remote X
servers to gain privileges via vectors involving image type and geometry,
which triggers out-of-bounds read operations.

References
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-7942
Upstream patch
https://cgit.freedesktop.org/xorg/lib/libX11/commit/?id=8ea762f94f4c942d898fdeb590a1630c83235c17

Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
8 years agolibXrandr: fix for CVE-2016-7947 and CVE-2016-7948
Sona Sarmadi [Fri, 27 Jan 2017 11:59:35 +0000 (12:59 +0100)]
libXrandr: fix for CVE-2016-7947 and CVE-2016-7948

CVE-2016-7947
Insufficient validation of server responses result in Integer overflows

CVE-2016-7948
Insufficient validation of server responses result in various data mishandlings

References:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-7947
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-7948
https://lists.x.org/archives/xorg-announce/2016-October/002720.html

Upstream patch for both CVEs:
https://cgit.freedesktop.org/xorg/lib/libXrandr/commit/?id=a0df3e1c7728205e5c7650b2e6dce684139254a6

Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
8 years agolibxrender: CVE-2016-7949
Sona Sarmadi [Fri, 27 Jan 2017 07:26:12 +0000 (08:26 +0100)]
libxrender: CVE-2016-7949

Insufficient validation of server responses results
in overflow of previously reserved memory

Upstream patch:
https://cgit.freedesktop.org/xorg/lib/libXrender/commit/?id=9362c7ddd1af3b168953d0737877bc52d79c94f4

External References:
https://lists.x.org/archives/xorg-announce/2016-October/002720.html
https://bugzilla.redhat.com/show_bug.cgi?id=CVE-2016-7949

Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
8 years agoclasses: Fix alternatives and rc.d ordering
David Vincent [Wed, 25 Jan 2017 13:03:06 +0000 (15:03 +0200)]
classes: Fix alternatives and rc.d ordering

When using an alternative as an initscript, the ordering between
update-rc.d and update-alternatives tasks during prerm and postinst
tasks must always be the following in order to work:
  * prerm:
    - stop daemon
    - remove alternative

  * postinst:
    - add alternative
    - start daemon

This patchset adds comments to the scripts generated by both classes and
organize the generated sections based on those comments.

[YOCTO #10433]

Signed-off-by: David Vincent <freesilicon@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(cherry picked from commit aa87b1a4dcc14e4dfe719b6c55045c5662bc59c2)

8 years agonspr: set correct version in pkg-config file
Andreas Müller [Wed, 14 Dec 2016 22:28:24 +0000 (23:28 +0100)]
nspr: set correct version in pkg-config file

while we are at it clean up sed

(From OE-Core rev: 69e9f7ec8ba8649784a27e9c7dc7a6987a53c22e)

(From OE-Core rev: cb4f5ecf32006192baaf41218a3d7b43ca1c2951)

Signed-off-by: Andreas Müller <schnitzeltony@googlemail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agoattr/ea-acl: pass --disable-gettext when USE_NLS=no
Denys Dmytriyenko [Wed, 11 Jan 2017 00:21:28 +0000 (19:21 -0500)]
attr/ea-acl: pass --disable-gettext when USE_NLS=no

Fixes following error in configure:
  FATAL ERROR: msgfmt does not seem to be installed.
  attr cannot be built without a working gettext installation.

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(From OE-Core rev: d044fad8a0ac5d57deb88b25106f3a39cb7c1636)

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-dtb: strip DTB extension properly in postinst/postrm
Denys Dmytriyenko [Tue, 10 Jan 2017 23:11:39 +0000 (18:11 -0500)]
linux-dtb: strip DTB extension properly in postinst/postrm

The use of awk -F "." in do_install/do_deploy to strip filename extension
was deprecated long time ago in 72980d5bb465f0640ed451d1ebb9c5d2a210ad0c.

Make a similar change in postinst/postrm to properly use basename command.

Otherwise DTB files that contain dots in the name result in broken symlinks
that point to non-existent truncated files.

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
(From OE-Core rev: 40c2addf0f0ee16b1c1334cf00f1490ffeaac475)

Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolibxtst: 1.2.2 -> 1.2.3
Sona Sarmadi [Tue, 10 Jan 2017 11:10:51 +0000 (12:10 +0100)]
libxtst: 1.2.2 -> 1.2.3

Upgrade libxtst from 1.2.2 to 1.2.3 to address:
Out of Bounds Write Denial of Service Vulnerability, CVE-2016-7951

References:
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-7951
https://cgit.freedesktop.org/xorg/lib/libXtst/commit/?id=9556ad67af3129ec4a7a4f4b54a0d59701beeae3

Signed-off-by: Sona Sarmadi <sona.sarmadi@enea.com>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agobinutils-2.27.inc: Fix alignment frags for aarch64
Manjukumar Matha [Sat, 17 Dec 2016 01:43:31 +0000 (17:43 -0800)]
binutils-2.27.inc: Fix alignment frags for aarch64

There was bug with alignment frags for aarch64 in binutils. This is fixed in
master of binutils. This patch backports the fix to binutils 2.27 version.

(From OE-Core rev: f6f87019073d4f3caa7766aca89faa6781690fba)

Signed-off-by: Manjukumar Matha <manjukumar.harthikote-matha@xilinx.com>
Signed-off-by: Nathan Rossi <nathan@nathanrossi.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agopopulate_sdk_ext: whitelist do_package tasks
Ed Bartosh [Wed, 4 Jan 2017 18:48:18 +0000 (20:48 +0200)]
populate_sdk_ext: whitelist do_package tasks

With enabled SSTATE_MIRRORS sstate code expects mirrors to
contain entries for all tasks, which is not the case for ext
installer as it uses reduced sstate cache.

Added do_package tasks to BB_SETSCENE_ENFORCE_WHITELIST to prevent
installer failing with ERROR: Sstate artifact unavailable

[YOCTO #10832]

(From OE-Core rev: 2ed46ada4b8e496493835e84b36f7e9c367f59d2)

Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agopopulate_sdk_ext: fix working with uninative sstate
Ed Bartosh [Wed, 4 Jan 2017 18:48:17 +0000 (20:48 +0200)]
populate_sdk_ext: fix working with uninative sstate

Mapped uninative sstate directories to make ext SDK installer to
use them when it's run on systems with gcc version different from
gcc version used to build installer.

[YOCTO #10832]

Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agozlib: update SRC_URI to fix fetching
Joshua Lock [Thu, 5 Jan 2017 16:34:23 +0000 (16:34 +0000)]
zlib: update SRC_URI to fix fetching

Upstream have removed the file from zlib.net as a new version has
been released, switch to fetching from the official sourceforge
mirror.

[YOCTO #10879]

(From OE-Core rev: bb99e4a620efd59556539c156cd98ea23aae74c8)

Signed-off-by: Joshua Lock <joshua.g.lock@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobuild-appliance-image: Update to morty head revision
Richard Purdie [Wed, 11 Jan 2017 17:26:48 +0000 (17:26 +0000)]
build-appliance-image: Update to morty head revision

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolibgcc-common: Don't apply symlinks for nativesdk
Richard Purdie [Wed, 21 Dec 2016 13:33:07 +0000 (13:33 +0000)]
libgcc-common: Don't apply symlinks for nativesdk

nativesdk-libgcc doesn't need a symlink into the target space and if we do this
sstate installation of the recipe can fail depending on whether it races with
the cross-canadian toolchains.

(From OE-Core rev: b2c1e1fe4221862e0dbf5d08960f0d0228e47c72)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake.conf: Add inclusion of BB_CURRENT_MC.conf after local.conf
Richard Purdie [Tue, 20 Dec 2016 19:05:01 +0000 (19:05 +0000)]
bitbake.conf: Add inclusion of BB_CURRENT_MC.conf after local.conf

People are strugling with multiconfig as the up front inclusion of the
configuration file doesn't do what people expect. The only way to meet
user expectations is to include the file immediately after local.conf.

We add BB_CURRENT_MC to bitbake so that the metadata can determine when
to include the extra configuration.

(From OE-Core rev: a6a3894fb2cb2097d2404b8b8cb2b85df595cfa9)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agouninative: Parameterise the use of STAGING_DIR
Richard Purdie [Wed, 21 Dec 2016 14:13:09 +0000 (14:13 +0000)]
uninative: Parameterise the use of STAGING_DIR

This means that a user can change TMPDIR in a multiconfig situation
and still only have one path to the uninative setup. Without this change
its not possile to make such a setup work.

(From OE-Core rev: 779422c5458f5f643b3a4a0dedaa4d9ad709367a)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoselftest: wic: fix test_qemu
Ed Bartosh [Fri, 2 Dec 2016 23:48:06 +0000 (01:48 +0200)]
selftest: wic: fix test_qemu

Setting WKS_FILE variable in qemux86-64 made wic test to
use wrong wks file to produce an image and resulted in
test_qemu failure.

Used conditional assignment in qemux86-64 and explicitly
set WKS_FILE in wic testing suite to make the suite to use
wic-image-minimal.wsk. This should fix test_qemu failure.

(From OE-Core rev: 3bca4d18c2712e3b154bacfb917f0a749ebaddeb)

Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agoutils: Always use datastore's PATH for host_gcc_version
Ross Burton [Thu, 15 Dec 2016 19:09:16 +0000 (19:09 +0000)]
utils: Always use datastore's PATH for host_gcc_version

BUILD_CC may reference something like ccache and expect this to come from
ccache-native, we at least have some selftests which assume this. Modify the
code to use PATH when runnig BUILD_CC to ensure the tests continue to work
as expected.

(From OE-Core rev: f3e753372baac43d0921186340cf260df056de20)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Hand applied and used d.getVar(True)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agovalgrind: make ld-XXX.so strlen intercept optional
Jackie Huang [Wed, 14 Dec 2016 02:26:13 +0000 (10:26 +0800)]
valgrind: make ld-XXX.so strlen intercept optional

Hack: Depending on how glibc was compiled (e.g. optimised
for size or built with _FORTIFY_SOURCE enabled) the strlen
symbol might not be found in ld-XXX.so. Therefore although
we should still try to intercept it, don't make it mandatory
to do so.

(From OE-Core rev: 84ec50e587e7464b260b1b189659b93b6dab0ef6)

Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agouninative: rebuild uninative for gcc 4.8 and 4.9
Ed Bartosh [Tue, 13 Dec 2016 23:55:04 +0000 (01:55 +0200)]
uninative: rebuild uninative for gcc 4.8 and 4.9

Some c++ libraries fail to build if uninative is built
with gcc 5.x and host gcc version is either 4.8 or 4.9.

The issue should be solved by making separate uninative sstate
directory structure sstate-cache/universal-<gcc version> for host gcc
versions 4.8 and 4.9. This causes rebuilds of uninative if host gcc
is either 4.8 or 4.9 and it doesn't match gcc version used to build
uninative.

[YOCTO #10441]

(From OE-Core rev: d36f41e5658bbbb6080ee833027879c119edf3e0)

Signed-off-by: Ed Bartosh <ed.bartosh@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolibxml2: Fix more NULL pointer derefs
Andrej Valek [Mon, 12 Dec 2016 13:20:21 +0000 (14:20 +0100)]
libxml2: Fix more NULL pointer derefs

The NULL pointer dereferencing could produced some
security problems.
This is a preventive security fix.

(From OE-Core rev: 8f3008114d5000a0865f50833db7c3a3f9808601)

Signed-off-by: Andrej Valek <andrej.valek@siemens.com>
Signed-off-by: Pascal Bach <pascal.bach@siemens.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolibxml2: fix CVE-2016-4658 Disallow namespace nodes in XPointer points and ranges
Andrej Valek [Mon, 12 Dec 2016 13:20:20 +0000 (14:20 +0100)]
libxml2: fix CVE-2016-4658 Disallow namespace nodes in XPointer points and ranges

Namespace nodes must be copied to avoid use-after-free errors.
But they don't necessarily have a physical representation in a
document, so simply disallow them in XPointer ranges.

(From OE-Core rev: 00e928bd1c2aed9caeaf9e411743805d2139a023)

Signed-off-by: Andrej Valek <andrej.valek@siemens.com>
Signed-off-by: Pascal Bach <pascal.bach@siemens.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolibxml2: Necessary changes before fixing CVE-2016-5131
Andrej Valek [Mon, 12 Dec 2016 13:20:19 +0000 (14:20 +0100)]
libxml2: Necessary changes before fixing CVE-2016-5131

xpath:
 - Check for errors after evaluating first operand.
 - Add sanity check for empty stack.
 - Include comparation in changes from xmlXPathCmpNodesExt to xmlXPathCmpNodes

(From OE-Core rev: 96ef568f75dded56a2123b63dcc8b443f796afe0)

Signed-off-by: Andrej Valek <andrej.valek@siemens.com>
Signed-off-by: Pascal Bach <pascal.bach@siemens.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolib/oe/rootfs: fix log_check warnings being printed twice with RPM packaging
Paul Eggleton [Mon, 12 Dec 2016 22:05:05 +0000 (11:05 +1300)]
lib/oe/rootfs: fix log_check warnings being printed twice with RPM packaging

We were calling _log_check() in the RPM-specific rootfs class as well as
in the base class; this is unnecessary and resulted in any errors/warnings
generated during the actual package installation time triggering two warnings
instead of one. Drop the call from RpmRootfs._create() to fix this.

(From OE-Core rev: 541c56d755ba0354297673e857628026ad9e4df2)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolibtiff: Update to 4.0.7
Armin Kuster [Sat, 10 Dec 2016 17:38:43 +0000 (09:38 -0800)]
libtiff: Update to 4.0.7

Major changes:
The libtiff tools bmp2tiff, gif2tiff, ras2tiff, sgi2tiff, sgisv, and ycbcr are completely removed from the distribution, used for demos.

CVEs fixed:
CVE-2016-9297
CVE-2016-9448
CVE-2016-9273
CVE-2014-8127
CVE-2016-3658
CVE-2016-5875
CVE-2016-5652
CVE-2016-3632

plus more that are not identified in the changelog.

removed patches integrated into update.
more info: http://libtiff.maptools.org/v4.0.7.html

(From OE-Core rev: 9945cbccc4c737c84ad441773061acbf90c7baed)

Signed-off-by: Armin Kuster <akuster808@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agokernel-yocto: explicitly trap subcommand errors
Bruce Ashfield [Fri, 9 Dec 2016 19:28:00 +0000 (14:28 -0500)]
kernel-yocto: explicitly trap subcommand errors

To trap errors and halt processing, do_kernel_metadata was recently
switched to exit on any non zero return code. While the concept is
sound, there are subcommands that have legitimate non-zero return
codes.

Instead of removing set +e, we'll explicitly check the return code
of the commands that can error, and throw a bbfatal to alert the
user.

(From OE-Core rev: a4705e62d0973c290011fc0d250501d358b659e8)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yocto/4.8: update to -rt7
Bruce Ashfield [Fri, 9 Dec 2016 19:27:59 +0000 (14:27 -0500)]
linux-yocto/4.8: update to -rt7

Updating to the latest 4.8-rt

(From OE-Core rev: 9f4565a308be55c1bf11706041c0565d48bda4f4)

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agoglibc: Enable backtrace from abort on ARM
Yuanjie Huang [Fri, 9 Dec 2016 17:49:34 +0000 (11:49 -0600)]
glibc: Enable backtrace from abort on ARM

ARM stack frames for abort and raise were limited to the the actual
abort and raise call, such as:

Obtained 4 stack frames.
./test-app(print_trace+0x1c) [0x10a08]
./test-app() [0x10b3c]
/lib/libc.so.6(__default_sa_restorer+0) [0x4adae1e0]
/lib/libc.so.6(gsignal+0xa0) [0x4adacf74]

This is not terribly useful when trying to figure out what function
may have called called the abort, especially when using pthreads.

After the change the trace would now look like:

Obtained 8 stack frames.
./test-app(print_trace+0x1c) [0x10a08]
./test-app() [0x10b3c]
/lib/libc.so.6(__default_sa_restorer+0) [0x4befe1e0]
/lib/libc.so.6(gsignal+0xa0) [0x4befcf74]
/lib/libc.so.6(abort+0x134) [0x4befe358]
./test-app(dummy_function+0x50) [0x10adc]
./test-app(main+0xd4) [0x10c24]
/lib/libc.so.6(__libc_start_main+0x114) [0x4bee7a58]

(From OE-Core rev: 93bf8713d8e13c278543baea94fb8dad0cb80e49)

Signed-off-by: Yuanjie Huang <yuanjie.huang@windriver.com>
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agoUse weak assignment for SERIAL_CONSOLES in qemu configuration files
Chen Qi [Fri, 9 Dec 2016 07:43:03 +0000 (15:43 +0800)]
Use weak assignment for SERIAL_CONSOLES in qemu configuration files

Use weak assignment for SERIAL_CONSOLES in qemu configuration files so that
the value could serve as a default value and could be easily overridden in
configuration files like local.conf.

When using the default value for SERIAL_CONSOLES in qemux86-64,we would have
annoying messages on console complaining about respawning getty on ttyS1.
Although the value is set by purpose, at least we need to provide an easy way
to override it.

(From OE-Core rev: 5f060b66162c41a295995947b918253450870117)

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agoarchiver: don't change directory when generating tarball
Ross Burton [Thu, 8 Dec 2016 21:29:57 +0000 (21:29 +0000)]
archiver: don't change directory when generating tarball

There's no need to chdir() as tarfile.add() can be told what name to use in the
archive.

(From OE-Core rev: d0b282bce34db44dde4dd7f53a64dfaafe6789de)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agogrub2: fix some quirks and div by zero
Awais Belal [Thu, 8 Dec 2016 14:09:39 +0000 (19:09 +0500)]
grub2: fix some quirks and div by zero

Rather than erroring out on a single attempt while
terminating EFI services, make a few retries because
such quirks are found in a few implementations.
Also fix a div by zero issue in the same framework
which causes an infinite reboot on the target.
Both patches included here are backports.

(From OE-Core rev: 5e6ac806bd9b8bf885ef1e88484e91e4cdaaa69a)

Signed-off-by: Awais Belal <awais_belal@mentor.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agowic: Create a logical partition only when it is really mandatory
Alessio Igor Bogani [Wed, 7 Dec 2016 14:00:57 +0000 (15:00 +0100)]
wic: Create a logical partition only when it is really mandatory

Don't worth bother with logical partition on MBR partition type (aka
msdos) if disk image generated by wic should have 4 partitions.

(From OE-Core rev: 36a558fbdc96094626e7de1a3510691e30885368)

Signed-off-by: Alessio Igor Bogani <alessio.bogani@elettra.eu>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolibarchive: fix ALTERNATIVE_PRIORITY to avoid conflict
Chen Qi [Wed, 7 Dec 2016 01:45:37 +0000 (09:45 +0800)]
libarchive: fix ALTERNATIVE_PRIORITY to avoid conflict

'tar' utility from tar and bsdtar has the same alternative priority.
'cpio' utility from cpio and bsdcpio has the same alternative priority.

Lower the ALTERNATIVE_PRIORITY to avoid conflict.

(From OE-Core rev: 9a59ff628771b586666999d44923968a6bc58956)

Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agocve-check: allow recipes to override the product name
Ross Burton [Wed, 7 Dec 2016 11:25:53 +0000 (11:25 +0000)]
cve-check: allow recipes to override the product name

Add a new variable CVE_PRODUCT for the product name to look up in the NVD
database.  Default this to BPN, but allow recipes such as tiff (which is libtiff
in NVD) to override it.

(From OE-Core rev: ba330051570a4c991885ee726cb187e0c911bd4f)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agocurl: set CVE_PRODUCT
Ross Burton [Wed, 7 Dec 2016 16:41:40 +0000 (16:41 +0000)]
curl: set CVE_PRODUCT

This is 'libcurl' in NVD.

(From OE-Core rev: f5381da49ac781ef017a1b9816c00b512ca9c7c2)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agotiff: set CVE_PRODUCT
Ross Burton [Thu, 8 Dec 2016 10:42:25 +0000 (10:42 +0000)]
tiff: set CVE_PRODUCT

This is 'libtiff' in NVD.

(From OE-Core rev: 0c8d1523f3ad0ada2d1b8f9abffbc2b898a744ca)

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agotargetloader.py: drop test for ClassType
Juro Bystricky [Sat, 10 Dec 2016 17:21:45 +0000 (09:21 -0800)]
targetloader.py: drop test for ClassType

ClassType was removed from python3.
The code testing for ClassType kept throwing AttributeError exceptions:

    module 'types' has no attribute 'ClassType'

The exceptions prevented loading of any dynamically resolved target
controllers.

(From OE-Core rev: d62f18c39bc0ed3b0f5ac8465b393c15f2143ecf)

Signed-off-by: Juro Bystricky <juro.bystricky@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yocto/4.x: CVE-2016-8655
Bruce Ashfield [Thu, 8 Dec 2016 17:58:55 +0000 (12:58 -0500)]
linux-yocto/4.x: CVE-2016-8655

Backporting upstream commit 84ac726023 [packet: fix race condition in
packet_set_ring] to address CVE-2016-8655

(From OE-Core rev: db7799747b263507427e325638353142ae79403c)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yocto-rt/4.4: update to -rt43
Bruce Ashfield [Thu, 8 Dec 2016 17:58:54 +0000 (12:58 -0500)]
linux-yocto-rt/4.4: update to -rt43

After the update to 4.4.36 -rt no longer builds. The fixes for the
issues are found in the v4.4-rt43 release.

(From OE-Core rev: 262fff10ba5cdedbee9ba9ecf00f98dc9159477c)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agokern-tools: ensure that no shared directories are used
Bruce Ashfield [Thu, 8 Dec 2016 17:58:53 +0000 (12:58 -0500)]
kern-tools: ensure that no shared directories are used

We need to avoid using shared/common directories for any files that are
part of specific build, since permissions issues in multi user
environments will cause issues.

Integrating the following commit to solve the issue:

   scc: move unused patch queue under output dir

(From OE-Core rev: cad65cc0eef2e06cb5ae08062ffae7a4d43a51ad)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yocto/4.8: update to v4.8.12
Bruce Ashfield [Thu, 8 Dec 2016 17:58:52 +0000 (12:58 -0500)]
linux-yocto/4.8: update to v4.8.12

shortlog of the changes follow:

   356ccf6d2b0c Linux 4.8.12
   f81c90012f2f scsi: mpt3sas: Unblock device after controller reset
   3de3eebb10fd flow_dissect: call init_default_flow_dissectors() earlier
   7838fbe25a95 mm, oom: stop pre-mature high-order OOM killer invocations
   374ff835e2b6 can: bcm: fix support for CAN FD frames
   7ed8d94b99b3 powerpc/boot: Fix the early OPAL console wrappers
   7cbe9568fa04 powerpc/mm: Fixup kernel read only mapping
   30988ea31783 powerpc: Set missing wakeup bit in LPCR on POWER9
   8d248df485f2 device-dax: fail all private mapping attempts
   f87a483126b5 device-dax: check devm_nsio_enable() return value
   5b08489eba14 xc2028: Fix use-after-free bug properly
   9030deb21f29 X.509: Fix double free in x509_cert_parse() [ver #3]
   0257b7e20c07 mpi: Fix NULL ptr dereference in mpi_powm() [ver #3]
   bdab996a7a9e mmc: sdhci-of-esdhc: fixup PRESENT_STATE read
   0a80bef18a9f parisc: Also flush data TLB in flush_icache_page_asm
   09e35a83cdc9 parisc: Fix race in pci-dma.c
   3cb2bc6fdc54 parisc: Switch to generic sched_clock implementation
   113b60efa37c parisc: Fix races in parisc_setup_cache_timing()
   b2f64572db2e thermal/powerclamp: add back module device table
   6dfe1643b56a perf/x86/intel: Cure bogus unwind from PEBS entries
   403f47ddbf6b perf/core: Fix address filter parser
   c5795c5f682d x86/traps: Ignore high word of regs->cs in early_fixup_exception()
   885bad1e5f32 x86/fpu: Fix invalid FPU ptrace state after execve()
   1a42cd56229b NFSv4.x: hide array-bounds warning
   154c665b2b02 apparmor: fix change_hat not finding hat after policy replacement
   9c15a98542f0 cfg80211: limit scan results cache size
   f204d21ee0b7 tile: avoid using clocksource_cyc2ns with absolute cycle count
   c707019bef62 scsi: mpt3sas: Fix secure erase premature termination
   b33387f6a68d Fix USB CB/CBI storage devices with CONFIG_VMAP_STACK=y
   89a28f3a3c77 USB: serial: ftdi_sio: add support for TI CC3200 LaunchPad
   6476f6530187 USB: serial: cp210x: add ID for the Zone DPMX
   7445ffb75ed5 usb: chipidea: move the lock initialization to core file
   32fe669c8634 KVM: x86: check for pic and ioapic presence before use
   c02d13809180 KVM: x86: fix out-of-bounds accesses of rtc_eoi map
   39b653013527 KVM: x86: drop error recovery in em_jmp_far and em_ret_far
   3c22c81377d4 KVM: x86: fix out-of-bounds access in lapic
   eb060c1a03f7 iommu/vt-d: Fix IOMMU lookup for SR-IOV Virtual Functions
   8b23f16343f6 iommu/vt-d: Fix PASID table allocation
   36bd5bfe4349 Linux 4.8.11
   a9a0027757f8 gpio: pca953x: Fix corruption of other gpios in set_multiple.
   88fce76cf590 gpio: pca953x: Move memcpy into mutex lock for set multiple
   9a7b80894951 netfilter: nft_dynset: fix element timeout for HZ != 1000
   fd4251fa4d09 IB/cm: Mark stale CM id's whenever the mad agent was unregistered
   8ce92be782f8 IB/uverbs: Fix leak of XRC target QPs
   160149299f71 IB/hfi1: Remove incorrect IS_ERR check
   46e47543175b IB/core: Avoid unsigned int overflow in sg_alloc_table
   82c377d05311 IB/mlx5: Fix NULL pointer dereference on debug print
   554e4b69f1bb IB/mlx5: Fix fatal error dispatching
   32091ee63d17 IB/mlx5: Fix memory leak in query device
   21822a5bbddd IB/mlx5: Use cache line size to select CQE stride
   ad886a0e9035 IB/mlx5: Validate requested RQT size
   76a93a6f8dd7 IB/mlx4: Fix create CQ error flow
   4234e6a06f8b IB/mlx4: Check gid_index return value
   fc9275365065 IB/rxe: Clear queue buffer when modifying QP to reset
   3c87b4a8f6d5 IB/rxe: Fix handling of erroneous WR
   c4605a0e9605 IB/rxe: Fix kernel panic in UDP tunnel with GRO and RX checksum
   04d016249998 IB/rxe: Update qp state for user query
   bac1543b764c perf hists: Fix column length on --hierarchy
   4cba876de64d PM / sleep: don't suspend parent when async child suspend_{noirq, late} fails
   2e2c8f0e7502 PM / sleep: fix device reference leak in test_suspend
   fd669bf2a099 uwb: fix device reference leaks
   ab17baeaea91 sunrpc: svc_age_temp_xprts_now should not call setsockopt non-tcp transports
   8c7ea73e3a9b mfd: core: Fix device reference leak in mfd_clone_cell
   545ea4003db9 iwlwifi: mvm: wake the wait queue when the RX sync counter is zero
   8bfaf856695b iwlwifi: mvm: fix d3_test with unified D0/D3 images
   8cdfd32398b5 iwlwifi: mvm: fix netdetect starting/stopping for unified images
   6e3324dc039c iwlwifi: pcie: mark command queue lock with separate lockdep class
   b822907865cc iwlwifi: pcie: fix SPLC structure parsing
   edb60ee4bc65 rtc: omap: Fix selecting external osc
   5f95e68daae3 clk: imx: fix integer overflow in AV PLL round rate
   fec43900c9b7 clk: mmp: mmp2: fix return value check in mmp2_clk_init()
   63465eaf7b2e clk: mmp: pxa168: fix return value check in pxa168_clk_init()
   a42bbd552ded clk: mmp: pxa910: fix return value check in pxa910_clk_init()
   bb6c9ec826be virtio-net: drop legacy features in virtio 1 mode
   899f5426eebf drm/i915: Assume non-DP++ port if dvo_port is HDMI and there's no AUX ch specified in the VBT
   f6920e506992 drm/i915: Refresh that status of MST capable connectors in ->detect()
   56a02a5f60ea drm/amdgpu: Attach exclusive fence to prime exported bo's. (v5)
   3eac4767dd7e powerpc/64: Fix setting of AIL in hypervisor mode
   fc312878ffe3 crypto: caam - do not register AES-XTS mode on LP units
   54f28973e8a5 ARM: dts: imx53-qsb: Fix regulator constraints
   7001b98b9dce ext4: sanity check the block and cluster size at mount time
   56df604296c2 kbuild: Steal gcc's pie from the very beginning
   e14754cb8f0c x86/kexec: add -fno-PIE
   ad4e2f324ade scripts/has-stack-protector: add -fno-PIE
   1c7727d0bca0 kbuild: add -fno-PIE
   20bcbe246933 Disable the __builtin_return_address() warning globally after all
   504b60516ba7 i2c: i2c-mux-pca954x: fix deselect enabling for device-tree
   45244660281c i2c: mux: fix up dependencies
   ce97f5012b6d ALSA: hda - Fix mic regression by ASRock mobo fixup
   23e14ee1abcf ALSA: hda - add a new condition to check if it is thinkpad
   85fcb62a58b2 ALSA: usb-audio: Fix use-after-free of usb_device at disconnect
   e902f10da218 gpio: do not double-check direction on sleeping chips
   b21b327d96bc can: bcm: fix warning in bcm_connect/proc_register
   50e6cd2feff7 mfd: stmpe: Fix RESET regression on STMPE2401
   e015527c6fbd mfd: intel-lpss: Do not put device in reset state on suspend
   8b4d44f46bf2 IB/hfi1: Fix rnr_timer addition
   06eac15f11e0 IB/rdmavt: rdmavt can handle non aligned page maps
   ca720a2b11b4 fuse: fix fuse_write_end() if zero bytes were copied
   de58c50e84e2 genirq: Use irq type from irqdata instead of irqdesc
   8844024c8839 ftrace: Add more checks for FTRACE_FL_DISABLED in processing ip records
   c5d20ce0b459 ftrace: Ignore FTRACE_FL_DISABLED while walking dyn_ftrace records
   f271087fb2ea KVM: arm64: Fix the issues when guest PMCCFILTR is configured
   8d8b37e242de arm64: KVM: pmu: Fix AArch32 cycle counter access
   1cb9b2489e77 KVM: Disable irq while unregistering user notifier
   23555ca21394 KVM: x86: fix missed SRCU usage in kvm_lapic_set_vapic_addr
   f8c74cf95655 x86/cpu/AMD: Fix cpu_llc_id for AMD Fam17h systems

(From OE-Core rev: beb9adb4763a1bbed182503371921e676ccd4ae2)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yocto/4.4: update to v4.4.36
Bruce Ashfield [Thu, 8 Dec 2016 17:58:51 +0000 (12:58 -0500)]
linux-yocto/4.4: update to v4.4.36

shortlog of the changes follow:

   87c6c6ef5b17 Linux 4.4.36
   6eddf5c993dd scsi: mpt3sas: Unblock device after controller reset
   c178e4809df7 flow_dissect: call init_default_flow_dissectors() earlier
   0b7860d6e88c mei: fix return value on disconnection
   bab2f72f70ea mei: me: fix place for kaby point device ids.
   4f13967ecd20 mei: me: disable driver on SPT SPS firmware
   e1049372d7a7 drm/radeon: Ensure vblank interrupt is enabled on DPMS transition to on
   249090830942 mpi: Fix NULL ptr dereference in mpi_powm() [ver #3]
   cd4235a794c0 parisc: Also flush data TLB in flush_icache_page_asm
   7a1ab6a2bf3a parisc: Fix race in pci-dma.c
   e541fd815db9 parisc: Fix races in parisc_setup_cache_timing()
   0c0ddbf7efec NFSv4.x: hide array-bounds warning
   be79d7fa43f9 apparmor: fix change_hat not finding hat after policy replacement
   8316338a201b cfg80211: limit scan results cache size
   4df31626fc08 tile: avoid using clocksource_cyc2ns with absolute cycle count
   ffffc1ed47e7 scsi: mpt3sas: Fix secure erase premature termination
   ab0867dd8bc8 Fix USB CB/CBI storage devices with CONFIG_VMAP_STACK=y
   f3f95f177269 USB: serial: ftdi_sio: add support for TI CC3200 LaunchPad
   1f36db0b397f USB: serial: cp210x: add ID for the Zone DPMX
   55d061bf9d23 usb: chipidea: move the lock initialization to core file
   341f9730c29b KVM: x86: check for pic and ioapic presence before use
   b7f9404d1b48 KVM: x86: drop error recovery in em_jmp_far and em_ret_far
   c091bbddbc5e iommu/vt-d: Fix IOMMU lookup for SR-IOV Virtual Functions
   d88a1bd00cfa iommu/vt-d: Fix PASID table allocation
   0a5b451a53d4 Linux 4.4.35
   1ba7fafae3c2 netfilter: nft_dynset: fix element timeout for HZ != 1000
   698a8dddb8a5 IB/cm: Mark stale CM id's whenever the mad agent was unregistered
   2b026a265a9a IB/uverbs: Fix leak of XRC target QPs
   eba83a85caba IB/core: Avoid unsigned int overflow in sg_alloc_table
   c524185c81d4 IB/mlx5: Fix fatal error dispatching
   734039913373 IB/mlx5: Use cache line size to select CQE stride
   7cf5b7882184 IB/mlx4: Fix create CQ error flow
   41664d7077e9 IB/mlx4: Check gid_index return value
   04c0800c73b2 PM / sleep: don't suspend parent when async child suspend_{noirq, late} fails
   469fcbcb84d8 PM / sleep: fix device reference leak in test_suspend
   680bc27065b9 uwb: fix device reference leaks
   dd214a159de6 mfd: core: Fix device reference leak in mfd_clone_cell
   d2adb5ebec61 iwlwifi: pcie: fix SPLC structure parsing
   1f995573bac8 rtc: omap: Fix selecting external osc
   c8aa3e98c1a8 clk: mmp: mmp2: fix return value check in mmp2_clk_init()
   96576127ffb7 clk: mmp: pxa168: fix return value check in pxa168_clk_init()
   a4709b4582ac clk: mmp: pxa910: fix return value check in pxa910_clk_init()
   d039fc37ec7f drm/amdgpu: Attach exclusive fence to prime exported bo's. (v5)
   96f10a6239a2 crypto: caam - do not register AES-XTS mode on LP units
   454cf79b05c5 ext4: sanity check the block and cluster size at mount time
   147117cf23c0 kbuild: Steal gcc's pie from the very beginning
   e543f094a38a x86/kexec: add -fno-PIE
   3a868dde1824 scripts/has-stack-protector: add -fno-PIE
   f740b5cc39dd kbuild: add -fno-PIE
   936d157fa7ae i2c: mux: fix up dependencies
   4e583b89add6 can: bcm: fix warning in bcm_connect/proc_register
   0c6e0db9686b mfd: intel-lpss: Do not put device in reset state on suspend
   b7321bcc8b1c fuse: fix fuse_write_end() if zero bytes were copied
   d4a774fdb92f KVM: Disable irq while unregistering user notifier
   b689e86c9a8f KVM: x86: fix missed SRCU usage in kvm_lapic_set_vapic_addr
   aea9d760b8ba x86/cpu/AMD: Fix cpu_llc_id for AMD Fam17h systems
   4eb9a8100248 Linux 4.4.34
   b4bbdcef7d90 sparc64: Delete now unused user copy fixup functions.
   cb85910b0d45 sparc64: Delete now unused user copy assembler helpers.
   1c7e17b1c4d6 sparc64: Convert U3copy_{from,to}_user to accurate exception reporting.
   7181969338f8 sparc64: Convert NG2copy_{from,to}_user to accurate exception reporting.
   bfc8be659309 sparc64: Convert NGcopy_{from,to}_user to accurate exception reporting.
   dc3a7a7d2c85 sparc64: Convert NG4copy_{from,to}_user to accurate exception reporting.
   1731d90d8a55 sparc64: Convert U1copy_{from,to}_user to accurate exception reporting.
   8a444c770f6e sparc64: Convert GENcopy_{from,to}_user to accurate exception reporting.
   c718e917b3d3 sparc64: Convert copy_in_user to accurate exception reporting.
   dd8a78b2b6ad sparc64: Prepare to move to more saner user copy exception handling.
   756723ad553d sparc64: Delete __ret_efault.
   f5a69ff7486e sparc64: Handle extremely large kernel TLB range flushes more gracefully.
   d36a1ac49d24 sparc64: Fix illegal relative branches in hypervisor patched TLB cross-call code.
   5d8eb954763d sparc64: Fix instruction count in comment for __hypervisor_flush_tlb_pending.
   217f829ae967 sparc64: Fix illegal relative branches in hypervisor patched TLB code.
   2ba06323db41 sparc64: Handle extremely large kernel TSB range flushes sanely.
   75931800733c sparc: Handle negative offsets in arch_jump_label_transform
   8fd11efa2140 sparc64 mm: Fix base TSB sizing when hugetlb pages are used
   4e90b6880135 sparc: serial: sunhv: fix a double lock bug
   a395f7a66eec sparc: Don't leak context bits into thread->fault_address
   4e772c53ab98 tty: Prevent ldisc drivers from re-using stale tty fields
   225a24ae9733 tcp: take care of truncations done by sk_filter()
   ae9e052a58ef ipv4: use new_gw for redirect neigh lookup
   5c67f9477bb2 net: __skb_flow_dissect() must cap its return value
   b67ed647d135 sock: fix sendmmsg for partial sendmsg
   0650eeb4f187 fib_trie: Correct /proc/net/route off by one error
   3f8857a4971d sctp: assign assoc_id earlier in __sctp_connect
   65d29c185614 ipv6: dccp: add missing bind_conflict to dccp_ipv6_mapped
   99131760a885 ipv6: dccp: fix out of bound access in dccp_v6_err()
   a2df29ed840f dccp: fix out of bound access in dccp_v4_err()
   ad6d0a82016b dccp: do not send reset to already closed sockets
   69a5c7ca2e62 tcp: fix potential memory corruption
   8777977b22c4 ip6_tunnel: Clear IP6CB in ip6tunnel_xmit()
   c5bad811ca4d bgmac: stop clearing DMA receive control register right after it is set
   6e9ca1b61cc8 net: mangle zero checksum in skb_checksum_help()
   ac4c2cf6f57a net: clear sk_err_soft in sk_clone_lock()
   74e53a3a053f dctcp: avoid bogus doubling of cwnd after loss
   86429bd405de Linux 4.4.33
   21cc1a183a9d netfilter: fix namespace handling in nf_log_proc_dostring
   ee5dd6878886 btrfs: qgroup: Prevent qgroup->reserved from going subzero
   ae6d4df4a7a0 mmc: mxs: Initialize the spinlock prior to using it
   ae5b8dbfe6be ASoC: sun4i-codec: return error code instead of NULL when create_card fails
   125e84726d7c ACPI / APEI: Fix incorrect return value of ghes_proc()
   5cd2cd84d573 i40e: fix call of ndo_dflt_bridge_getlink()
   5be7e6b48b21 hwrng: core - Don't use a stack buffer in add_early_randomness()
   ba8580f6cf03 lib/genalloc.c: start search from start of chunk
   c048b6711ed6 mei: bus: fix received data size check in NFC fixup
   19426f065d1e iommu/vt-d: Fix dead-locks in disable_dmar_iommu() path
   f029e7b34f25 iommu/amd: Free domain id when free a domain of struct dma_ops_domain
   f5bb84163704 tty/serial: at91: fix hardware handshake on Atmel platforms
   49163391881a dmaengine: at_xdmac: fix spurious flag status for mem2mem transfers
   fff40ee4d224 drm/i915: Respect alternate_ddc_pin for all DDI ports
   d7b0055e5566 KVM: MIPS: Precalculate MMIO load resume PC
   d24587404922 scsi: mpt3sas: Fix for block device of raid exists even after deleting raid disk
   6e897d034d4c scsi: qla2xxx: Fix scsi scan hang triggered if adapter fails during init
   29cf142cefed iio: orientation: hid-sensor-rotation: Add PM function (fix non working driver)
   caff14e99c63 iio: hid-sensors: Increase the precision of scale to fix wrong reading interpretation.
   dbbc5e6bcf36 clk: qoriq: Don't allow CPU clocks higher than starting value
   2c5cdadeab5d toshiba-wmi: Fix loading the driver on non Toshiba laptops
   f523deedff82 drbd: Fix kernel_sendmsg() usage - potential NULL deref
   f1de32232db3 usb: gadget: u_ether: remove interrupt throttling
   0ab4186bf238 USB: cdc-acm: fix TIOCMIWAIT
   290ffd550e34 staging: nvec: remove managed resource from PS2 driver
   4aa859ea591e Revert "staging: nvec: ps2: change serio type to passthrough"
   729eb8d9f7b7 drivers: staging: nvec: remove bogus reset command for PS/2 interface
   6bba8c37ceed staging: iio: ad5933: avoid uninitialized variable in error case
   a98e483ca784 pinctrl: cherryview: Prevent possible interrupt storm on resume
   126d0a2fe878 pinctrl: cherryview: Serialize register access in suspend/resume
   c45bfaa9211e ARC: timer: rtc: implement read loop in "C" vs. inline asm
   3b21a0b468a3 s390/hypfs: Use get_free_page() instead of kmalloc to ensure page alignment
   e835220ed25f coredump: fix unfreezable coredumping task
   5c54f79ad234 swapfile: fix memory corruption via malformed swapfile
   2f7496c48a7b dib0700: fix nec repeat handling
   a04769029386 ASoC: cs4270: fix DAPM stream name mismatch
   f3155797b91c ALSA: info: Limit the proc text input size
   c1ea2b3c07c6 ALSA: info: Return error for invalid read/write

(From OE-Core rev: 7d8d71ba5bc2dd8884c3f43a8d71a6d399c4e2fe)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yoct/4.1: update to v4.1.36
Bruce Ashfield [Thu, 8 Dec 2016 17:58:50 +0000 (12:58 -0500)]
linux-yoct/4.1: update to v4.1.36

Shortlog of the changes:

   8576fa45c10e Linux 4.1.36
   39f99860357c kbuild: add -fno-PIE
   bf5d3d296abf firewire: net: fix fragmented datagram_size off-by-one
   c604dec3d5a6 firewire: net: guard against rx buffer overflows
   9fe6256c0020 parisc: Ensure consistent state when switching to kernel stack at syscall entry
   83a474ed7003 ovl: fsync after copy-up
   c0b309f18b01 virtio: console: Unlock vqs while freeing buffers
   4fe9ae4d3fcf md: be careful not lot leak internal curr_resync value into metadata. -- (all)
   e1e5cab9b43c md: sync sync_completed has correct value as recovery finishes.
   97d53c4d89e0 scsi: arcmsr: Send SYNCHRONIZE_CACHE command to firmware
   d207c6603fc6 scsi: scsi_debug: Fix memory leak if LBP enabled and module is unloaded
   169eb57c9e8d drm/radeon/si_dpm: workaround for SI kickers
   c1593e5dcc2a drm/dp/mst: Check peer device type before attempting EDID read
   e5c6bbbcc7be drm/dp/mst: add some defines for logical/physical ports
   dadd5803301f drm/dp/mst: Clear port->pdt when tearing down the i2c adapter
   a2d4bd9c99cc KVM: MIPS: Precalculate MMIO load resume PC
   b05ff0cb0dea KVM: MIPS: Make ERET handle ERL before EXL
   90a107c0b85b drm/radeon: drop register readback in cayman_cp_int_cntl_setup
   9a9a2373142a scsi: megaraid_sas: Fix data integrity failure for JBOD (passthrough) devices
   1b15bd739689 Revert "drm/radeon: fix DP link training issue with second 4K monitor"
   5bc028b10d04 powerpc: Convert cmp to cmpd in idle enter sequence
   00a4335dfaa5 drm/radeon/si_dpm: Limit clocks on HD86xx part
   469bc68e081c drm/radeon/si/dpm: add workaround for for Jet parts
   85a2af9f690a USB: serial: cp210x: fix tiocmget error handling
   0854ce492ec3 Input: i8042 - add XMG C504 to keyboard reset table
   0e9bbbdfe9f8 ALSA: hda - Merge RIRB_PRE_DELAY into CTX_WORKAROUND caps
   02426cf24407 mac80211: discard multicast and 4-addr A-MSDUs
   f065a553f36e ubifs: Fix regression in ubifs_readdir()
   5fdb4fad291f GenWQE: Fix bad page access during abort of resource allocation
   55a2f87cd564 mm/list_lru.c: avoid error-path NULL pointer deref
   52e7b6bd8378 btrfs: fix races on root_log_ctx lists
   9f2d48f0745f vt: clear selection before resizing
   93ec720c6f10 tty: limit terminal size to 4M chars
   ca963122cab2 ALSA: usb-audio: Add quirk for Syntek STK1160
   3dc8f1f0decb KEYS: Fix short sprintf buffer in /proc/keys show function
   1f7ff6f7910e hv: do not lose pending heartbeat vmbus packets
   9662d19bab68 mei: txe: don't clean an unprocessed interrupt cause.
   a64291d578dd ANDROID: binder: Clear binder and cookie when setting handle in flat binder struct
   11924cba404d ANDROID: binder: Add strong ref checks
   39709c84807f dm table: fix missing dm_put_target_type() in dm_table_add_target()
   c91812c8ef76 usb: increase ohci watchdog delay to 275 msec
   51b231bd1a95 arm64: KVM: Take S1 walks into account when determining S2 write faults
   bb21e0295064 arm/arm64: KVM: Handle out-of-RAM cache maintenance as a NOP
   7fbfac87fbf5 USB: serial: fix potential NULL-dereference at probe
   605a6960617f xhci: workaround for hosts missing CAS bit
   c23a6dce1147 xhci: add restart quirk for Intel Wildcatpoint PCH
   4c17f91bb370 kvm: x86: memset whole irq_eoi
   9c5e89d59d56 libxfs: clean up _calc_dquots_per_chunk
   d1375c3e4f2e target: Don't override EXTENDED_COPY xcopy_pt_cmd SCSI status code
   ac49951623e1 target: Re-add missing SCF_ACK_KREF assignment in v4.1.y
   eaf55207fe3b ubifs: Abort readdir upon error
   b9ce8700a2e3 UBIFS: Fix possible memory leak in ubifs_readdir()
   cad1ca2a9485 ubifs: Fix xattr_names length in exit paths
   dfeb7e4f30e9 arm64: percpu: rewrite ll/sc loops in assembly
   3c9c01324032 powerpc/mm: Prevent unlikely crash in copro_calculate_slb()
   c1f5e5f89b99 isofs: Do not return EACCES for unknown filesystems
   b15662c8e266 irqchip/gic-v3-its: Fix entry size mask for GITS_BASER
   af17243fc1b4 arm64: kernel: Init MDCR_EL2 even in the absence of a PMU
   c911b72eaff1 USB: serial: ftdi_sio: add support for Infineon TriBoard TC2X7
   59eabe8bb60f memstick: rtsx_usb_ms: Manage runtime PM when accessing the device
   86559fc191ca memstick: rtsx_usb_ms: Runtime resume the device when polling for cards
   a89634cf709b mmc: rtsx_usb_sdmmc: Handle runtime PM while changing the led
   e79c2f2f7b30 mmc: rtsx_usb_sdmmc: Avoid keeping the device runtime resumed when unused
   f125e7f2285c mmc: sdhci: cast unsigned int to unsigned long long to avoid unexpeted error
   517052446f36 usb: gadget: function: u_ether: don't starve tx request queue
   66f38bf99faa ceph: fix error handling in ceph_read_iter
   9e9ca450cd93 irqchip/gicv3: Handle loop timeout proper
   88b6c9f4146a drm/radeon: change vblank_time's calculation method to reduce computational error.
   dcf4c1144f3a jbd2: fix incorrect unlock on j_list_lock
   a3da255ead60 scsi: Fix use-after-free
   8df981058e4f mmc: core: Annotate cmd_hdr as __le32
   7ea3d2e9c6aa x86/mm: Expand the exception table logic to allow new handling options

(From OE-Core rev: 41ab5422d4dc62b6e21ccad975b2a22960d3a44b)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yocto/4.4/4.8: Fix remaining kernel_configcheck warnings in Intel BSPs
Bruce Ashfield [Thu, 8 Dec 2016 17:58:49 +0000 (12:58 -0500)]
linux-yocto/4.4/4.8: Fix remaining kernel_configcheck warnings in Intel BSPs

Integrating the following meta-data change to remove warnings during
kernel configuration checking:

  bec7e0ace0e5 leds.cfg: Remove CONFIG_LEDS_TRIGGER_CPU
  cc857e64ea34 bsp/intel-quark: Remove bosch-pressure-sensor-i2c
  3198ab25af74 features/i2c: Support I2C_MUX by default

(From OE-Core rev: 2cf5fae50a3b7f232ad667c6c60eaa3f2aafd149)

Signed-off-by: California Sullivan <california.l.sullivan@intel.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agolinux-yocto/4.8: aufs warning and ixgbe calltrace
Bruce Ashfield [Thu, 8 Dec 2016 17:58:48 +0000 (12:58 -0500)]
linux-yocto/4.8: aufs warning and ixgbe calltrace

Integrating the following commits to fix aufs/ixgbe issues:

   Author: Liwei Song <liwei.song@windriver.com>
   Date:   Mon Dec 5 00:50:16 2016 -0500

       ixgbe: initialize u64_stats_sync structures early at ixgbe_probe

       Fix the following CallTrace:
       INFO: trying to register non-static key.
       the code is fine but needs lockdep annotation.
       turning off the locking correctness validator.
       CPU: 71 PID: 1 Comm: swapper/0 Not tainted 4.8.8-WR9.0.0.1_standard #11
       Hardware name: Intel Corporation S2600WTT/S2600WTT,
       BIOS GRNDSDP1.86B.0036.R05.1407140519 07/14/2014
        00200086 00200086 eb5e1ab8 c144dd70 00000000 00000000 eb5e1af8 c10af89a
        c1d23de4 eb5e1af8 00000009 eb5d8600 eb5d8638 eb5e1af8 c10b14d8 00000009
        0000000a c1d32911 00000000 00000000 e44c826c eb5d8000 eb5e1b74 c10b214e
       Call Trace:
        [<c144dd70>] dump_stack+0x5f/0x8f
        [<c10af89a>] register_lock_class+0x25a/0x4c0
        [<c10b14d8>] ? check_irq_usage+0x88/0xc0
        [<c10b214e>] __lock_acquire+0x5e/0x17a0
        [<c1abdb9b>] ? _raw_spin_unlock_irqrestore+0x3b/0x70
        [<c10cf14a>] ? rcu_read_lock_sched_held+0x8a/0x90
        [<c10b3c5f>] lock_acquire+0x9f/0x1f0
        [<c1922dcf>] ? dev_get_stats+0x5f/0x110
        [<c176e6b3>] ixgbe_get_stats64+0x113/0x320
        [<c1922dcf>] ? dev_get_stats+0x5f/0x110
        [<c1922dcf>] dev_get_stats+0x5f/0x110
        [<c1ab5415>] rtnl_fill_stats+0x40/0x105
        [<c193dd45>] rtnl_fill_ifinfo+0x4c5/0xd20
        [<c11c5115>] ? __kmalloc_node_track_caller+0x1a5/0x410
        [<c1917487>] ? __kmalloc_reserve.isra.42+0x27/0x80
        [<c191754f>] ? __alloc_skb+0x6f/0x270
        [<c1942291>] rtmsg_ifinfo_build_skb+0x71/0xd0
        [<c194230a>] rtmsg_ifinfo.part.23+0x1a/0x50
        [<c1923dad>] ? call_netdevice_notifiers_info+0x2d/0x60
        [<c194236b>] rtmsg_ifinfo+0x2b/0x40
        [<c192f997>] register_netdevice+0x3d7/0x4d0
        [<c192faa7>] register_netdev+0x17/0x30
        [<c177b83d>] ixgbe_probe+0x118d/0x1610
        [<c1498202>] local_pci_probe+0x32/0x80
        [<c1498172>] ? pci_match_device+0xd2/0x100
        [<c14991e0>] pci_device_probe+0xc0/0x110
        [<c1652cc5>] driver_probe_device+0x1c5/0x280
        [<c1498172>] ? pci_match_device+0xd2/0x100
        [<c1652e09>] __driver_attach+0x89/0x90
        [<c1652d80>] ? driver_probe_device+0x280/0x280
        [<c165114f>] bus_for_each_dev+0x4f/0x80
        [<c165269e>] driver_attach+0x1e/0x20
        [<c1652d80>] ? driver_probe_device+0x280/0x280
        [<c1652317>] bus_add_driver+0x1a7/0x220
        [<c1653a79>] driver_register+0x59/0xe0
        [<c1f897b8>] ? igb_init_module+0x49/0x49
        [<c1497b2a>] __pci_register_driver+0x4a/0x50
        [<c1f8985d>] ixgbe_init_module+0xa5/0xc4
        [<c1000485>] do_one_initcall+0x35/0x150
        [<c107e818>] ? parameq+0x18/0x70
        [<c1f395d8>] ? repair_env_string+0x12/0x51
        [<c107ead0>] ? parse_args+0x260/0x3b0
        [<c1074f73>] ? __usermodehelper_set_disable_depth+0x43/0x50
        [<c1f39e90>] kernel_init_freeable+0x19b/0x267
        [<c1f395c6>] ? set_debug_rodata+0xf/0xf
        [<c10b1e7b>] ? trace_hardirqs_on+0xb/0x10
        [<c1abdc02>] ? _raw_spin_unlock_irq+0x32/0x50
        [<c1085f0b>] ? finish_task_switch+0xab/0x1f0
        [<c1085ec9>] ? finish_task_switch+0x69/0x1f0
        [<c1ab6a30>] kernel_init+0x10/0x110
        [<c108bd65>] ? schedule_tail+0x25/0x80
        [<c1abe422>] ret_from_kernel_thread+0xe/0x24
        [<c1ab6a20>] ? rest_init+0x130/0x130

       This CallTrace occurred on 32-bit kernel with CONFIG_PROVE_LOCKING
       enabled.

       This happens at ixgbe driver probe hardware stage, when comes to
       ixgbe_get_stats64, the seqcount/seqlock still not initialize, although
       this was initialize in TX/RX resources setup routin, but it was too late,
       then lockdep give this Warning.

       To fix this, move the u64_stats_init function to driver probe stage,
       which before we get the status of seqcount and after the RX/TX ring
       was finished init.

Signed-off-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
   Author: Kexin(Casey) Chen <Casey.Chen@windriver.com>
   Date:   Tue Nov 29 14:14:07 2016 -0800

       aufs: fix compile warning

       fs/aufs/debug.h:95:19: warning: comparison of constant '0'
       with boolean expression is always false [-Wbool-compare]
          if (unlikely((e) < 0)) \
                           ^

       fs/aufs/vdir.c:852:2: note: in expansion of macro 'AuTraceErr'
          AuTraceErr(!valid);
          ^~~~~~~~~~

       In expansion of AuTraceErr(!valid), comparison of (!valid)
       and constant '0' always passes unlikely(x) false. function
       'static int seek_vdir(struct file *file, struct dir_context *ctx)'
       is to find whether there is a valid vd_deblk following ctx->pos.
       return 1 means valid, 0 for not. Change to AuTraceErr(valid - 1)
       makes more sense.

Signed-off-by: Kexin(Casey) Chen <Casey.Chen@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
(From OE-Core rev: ae7d870954cf6a3608933335fedc425677aee0e3)

Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agodiffutils: do_configure: fix "Argument list too long"
Robert Yang [Mon, 5 Dec 2016 14:54:39 +0000 (06:54 -0800)]
diffutils: do_configure: fix "Argument list too long"

Fixed when len(TMPDIR) =  410:
aclocal: error: cannot open echo [snip]: Argument list too long

This is becuase it has a lot of m4 files, use relative path for them
can fix the problem.

It doesn't happen when MACHINE="qemux86", I think it is because
intel-x86-64 is longer than qemux86.

(From OE-Core rev: 5210ccd61ef52a191454a4587cfeb22079df746d)

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agokernel.bbclass: fix kernel_do_compile for KERNEL_IMAGETYPE = "vmlinux.gz" on mips
Andreas Oberritter [Wed, 30 Nov 2016 23:36:51 +0000 (00:36 +0100)]
kernel.bbclass: fix kernel_do_compile for KERNEL_IMAGETYPE = "vmlinux.gz" on mips

The target directory didn't exist.

(From OE-Core rev: bd62851dc236a0279c735b290782602e275de5c1)

Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agokernel.bbclass: do not copy bundled initramfs to /boot
Andreas Oberritter [Wed, 30 Nov 2016 23:36:50 +0000 (00:36 +0100)]
kernel.bbclass: do not copy bundled initramfs to /boot

These files cause warnings because they don't get packaged, now that they
don't land in kernel-vmlinux anymore.

(From OE-Core rev: a49569e3a7534779bbe3f01a0647fd076c95798d)

Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agokernel.bbclass: Avoid wildcards for kernel images
Andreas Oberritter [Wed, 30 Nov 2016 23:36:49 +0000 (00:36 +0100)]
kernel.bbclass: Avoid wildcards for kernel images

With multiple kernel images enabled starting with 'vmlinux', e.g.
vmlinux.gz and vmlinux.bin, all files landed inside the
kernel-vmlinux package.

On top of that, even initramfs images were included, e.g.
vmlinux.gz-initramfs-*.

(From OE-Core rev: b7f4133b44b740e8ac8e758b0d4a3ee32d326332)

Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agokernel.bbclass: Use real filenames in kernel packages
Andreas Oberritter [Wed, 30 Nov 2016 23:36:48 +0000 (00:36 +0100)]
kernel.bbclass: Use real filenames in kernel packages

When iterating over kernel image types to set up their packaging
variables, don't use make targets but the real names.

It was surprising if both vmlinux.bin and vmlinux.gz were enabled
and only the latter had its filename extension removed from the
package name.

(From OE-Core rev: aa189f183e10588f7e8d642f351bd9b8d69f3ea9)

Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agokernel.bbclass: allow uncompressed initramfs archives
Andreas Oberritter [Wed, 30 Nov 2016 23:36:47 +0000 (00:36 +0100)]
kernel.bbclass: allow uncompressed initramfs archives

The code failed to copy the initramfs in case it was a plain
cpio archive.

(From OE-Core rev: 7dbdb4ea91aa027866da2bd46c65fe65a25c848f)

Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agorecipetool: fix encoding-related errors creating python recipes
Paul Eggleton [Sun, 4 Dec 2016 22:11:44 +0000 (11:11 +1300)]
recipetool: fix encoding-related errors creating python recipes

Yet another instance of us expecting a string back from subprocess when
in Python 3 what you get back is bytes. Just decode the output within
run_command() so we avoid this everywhere.

(From OE-Core rev: 103faae78cdff5280c7b7cdb7ca01e0868d02ec9)

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agogrub2: enforce -no-pie if supported by compiler
Alexander Kanavin [Fri, 2 Dec 2016 19:14:07 +0000 (21:14 +0200)]
grub2: enforce -no-pie if supported by compiler

Recent distros are enabling -pie by default; in case of grub
we need to turn it off.

(From OE-Core rev: aaff6c99dde3f1058bb3c4b320f27753c6c992ad)

Signed-off-by: Alexander Kanavin <alexander.kanavin@linux.intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agotiff: Fix several CVE issues
Mingli Yu [Wed, 7 Dec 2016 08:01:13 +0000 (16:01 +0800)]
tiff: Fix several CVE issues

Fix CVE-2016-9533, CVE-2016-9534, CVE-2016-9536 and
CVE-2016-9537

External References:
http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2016-9533
http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2016-9534
http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2016-9536
http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2016-9537

Patch from:
https://github.com/vadz/libtiff/commit/83a4b92815ea04969d494416eaae3d4c6b338e4a#diff-c8b4b355f9b5c06d585b23138e1c185f

(From OE-Core rev: f75ecefee21ef89b147fff9afae01a6f09c93198)

Signed-off-by: Mingli Yu <Mingli.Yu@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agotiff: Security fix CVE-2016-9538
Mingli Yu [Wed, 7 Dec 2016 08:01:12 +0000 (16:01 +0800)]
tiff: Security fix CVE-2016-9538

* tools/tiffcrop.c: fix read of undefined buffer in
readContigStripsIntoBuffer() due to uint16 overflow.

External References:
http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2016-9538

Patch from:
https://github.com/vadz/libtiff/commit/43c0b81a818640429317c80fea1e66771e85024b#diff-c8b4b355f9b5c06d585b23138e1c185f

(From OE-Core rev: 9af5d5ea882c853e4cb15006f990d3814eeea9ae)

Signed-off-by: Mingli Yu <Mingli.Yu@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agotiff: Security fix CVE-2016-9535
Mingli Yu [Wed, 7 Dec 2016 08:01:11 +0000 (16:01 +0800)]
tiff: Security fix CVE-2016-9535

* libtiff/tif_predict.h, libtiff/tif_predict.c:
Replace assertions by runtime checks to avoid assertions in debug mode,
or buffer overflows in release mode. Can happen when dealing with
unusual tile size like YCbCr with subsampling.

External References:
http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2016-9535

Patch from:
https://github.com/vadz/libtiff/commit/3ca657a8793dd011bf869695d72ad31c779c3cc1
https://github.com/vadz/libtiff/commit/6a984bf7905c6621281588431f384e79d11a2e33

(From OE-Core rev: 61d3feb9cad9f61f6551b43f4f19bfa33cadd275)

Signed-off-by: Mingli Yu <Mingli.Yu@windriver.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agotiff: set CVE NAME
Ross Burton [Wed, 7 Dec 2016 11:26:01 +0000 (11:26 +0000)]
tiff: set CVE NAME

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agosubversion: Fix issues in LDFLAGS sed manipulation
Richard Purdie [Wed, 7 Dec 2016 13:54:18 +0000 (13:54 +0000)]
subversion: Fix issues in LDFLAGS sed manipulation

The existing sed expression can match expressions like
--sysroot=/some/path/xxx-linux/ which clearly isn't intended and
injects incorrect paths into LDFLAGS.

Fix this in the same way we address the problem in CFLAGS. This fixes corrupt
build paths and incorrect paths in .la files amongst other issues.

(From OE-Core rev: 9a8382422ddbb0972dc25b752204f4908bb9857c)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agoattr: Convert SSTATEPOSTINSTFUNCS to a do_install_append
Richard Purdie [Wed, 7 Dec 2016 12:07:31 +0000 (12:07 +0000)]
attr: Convert SSTATEPOSTINSTFUNCS to a do_install_append

A SSTATEPOSTINSTFUNCS function here is overkill, just do this in a
do_install_append_class-native and create relative symlinks rather
than absolute ones which would then have to be relocated.

(From OE-Core rev: 518e8d0216b0f42f574e42288804f553b9ff6f99)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>
8 years agostaging: Drop unused SYSROOT_LOCK
Richard Purdie [Wed, 7 Dec 2016 12:09:56 +0000 (12:09 +0000)]
staging: Drop unused SYSROOT_LOCK

This hasn't been used in years and isn't referenced anywhere, drop
the definition.

(From OE-Core rev: 68258f856f08f35813964e58ef761471e29373ad)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Armin Kuster <akuster808@gmail.com>