A DMA-MMIO reentrancy problem may lead to memory corruption bugs
like stack overflow or use-after-free.
Summary of the problem from Peter Maydell:
https://lore.kernel.org/qemu-devel/CAFEAcA_23vc7hE3iaM-JVA6W38LK4hJoWae5KcknhPRD5fPBZA@mail.gmail.com
Reference:
https://gitlab.com/qemu-project/qemu/-/issues/556
qemu.git$ git log --no-merges --oneline --grep CVE-2023-0330
b987718bbb hw/scsi/lsi53c895a: Fix reentrancy issues in the LSI controller (CVE-2023-0330)
a2e1753b80 memory: prevent dma-reentracy issues
Included second commit as well as commit log of a2e1753b80 says it
resolves CVE-2023-0330
(From OE-Core rev: 45ce9885351a2344737170e6e810dc67ab3e7ea9)
Signed-off-by: Vijay Anusuri <vanusuri@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
This CVE shouldn't have been filed as the "exploit" is described in the
documentation as how the library behaves.
(From OE-Core rev: b66a677b76c7f15eb5c426f8dc7ac42e1e2e3f40)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit c652f094d86c4efb7ff99accba63b8169493ab18)
Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
The runpath in the cc1 binary is:
Library runpath: [$ORIGIN/../../../recipe-sysroot-native/usr/lib:$ORIGIN/../../../recipe-sysroot-native/lib]
This does not match the actual location of the libraries, which would require:
Library runpath: [$ORIGIN/../../recipe-sysroot-native/usr/lib:$ORIGIN/../../recipe-sysroot-native/lib]
Prior to gcc 9.1 the recipe set B explicity with:
B = "${WORKDIR}/gcc-${PV}/build.${HOST_SYS}.${TARGET_SYS}"
and this build directory structure matches the runpath in cc1, so there is no issue.
This line was commented out in versions 9.1 through 11.3. The upgrade to 12.1 once
again uncommented this line.
As a result the runpath is incorrect in version 9.1 through 11.3 and cc1 defaults
to using host libraries.
This patch restores setting B as done in master and versions prior to 9.1
(From OE-Core rev: 43d5ebde6d609898064ea70c89a7eba002e5fd74)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
When building using an SDK, cmake complains that the target
architecture 'cortexa53-crypto' is unknown. The same build in bitbake
uses the target architecture 'aarch64'.
Set CMAKE_SYSTEM_PROCESSOR the same as for bitbake.
(From OE-Core rev: 7a7ef9d73affc23fa14712d56f1a40d0c46569cb)
Signed-off-by: Tom Hochstein <tom.hochstein@nxp.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit d32a6225eefce2073a1cd401034b5b4c68351bfe)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
License-Update: update year to 2023
30afa75ad8
Release Notes for 3.8.15:
Security content in this release
CVE-2022-40674: bundled libexpat was upgraded from 2.4.7 to 2.4.9 which
fixes a heap use-after-free vulnerability in function doContent
gh-97616: a fix for a possible buffer overflow in list *= int
gh-97612: a fix for possible shell injection in the example script
get-remote-certificate.py (this issue originally had a CVE assigned to
it, which its author withdrew)
gh-96577: a fix for a potential buffer overrun in msilib
https://www.python.org/downloads/release/python-3815/
Release Notes for 3.8.16:
Security content in this release
gh-98739: Updated bundled libexpat to 2.5.0 to fix CVE-2022-43680 (heap
use-after-free).
gh-98517: Port XKCP’s fix for the buffer overflows in SHA-3 to fix
CVE-2022-37454.
gh-98433: The IDNA codec decoder used on DNS hostnames by socket or
asyncio related name resolution functions no longer involves a quadratic
algorithm to fix CVE-2022-45061. This prevents a potential CPU denial of
service if an out-of-spec excessive length hostname involving
bidirectional characters were decoded. Some protocols such as urllib
http 3xx redirects potentially allow for an attacker to supply such a
name.
gh-68966: The deprecated mailcap module now refuses to inject unsafe
text (filenames, MIME types, parameters) into shell commands to address
CVE-2015-20107. Instead of using such text, it will warn and act as if a
match was not found (or for test commands, as if the test failed).
gh-100001: python -m http.server no longer allows terminal control
characters sent within a garbage request to be printed to the stderr
server log.
gh-87604: Avoid publishing list of active per-interpreter audit hooks
via the gc module.
https://www.python.org/downloads/release/python-3816/
Release Notes for 3.8.17:
Security content in this release
gh-103142: The version of OpenSSL used in Windows and Mac installers has
been upgraded to 1.1.1u to address CVE-2023-2650, CVE-2023-0465,
CVE-2023-0466, CVE-2023-0464, as well as CVE-2023-0286, CVE-2022-4303,
and CVE-2022-4303 fixed previously in 1.1.1t (gh-101727).
gh-102153: urllib.parse.urlsplit() now strips leading C0 control and
space characters following the specification for URLs defined by WHATWG
in response to CVE-2023-24329.
gh-99889: Fixed a security in flaw in uu.decode() that could allow for
directory traversal based on the input if no out_file was specified.
gh-104049: Do not expose the local on-disk location in directory indexes
produced by http.client.SimpleHTTPRequestHandler.
gh-103935: trace.__main__ now uses io.open_code() for files to be
executed instead of raw open().
gh-101283: subprocess.Popen now uses a safer approach to find cmd.exe
when launching with shell=True.
gh-102953: The extraction methods in tarfile, and
shutil.unpack_archive(), have a new filter argument that allows limiting
tar features than may be surprising or dangerous, such as creating files
outside the destination directory. See Extraction filters for details.
https://www.python.org/downloads/release/python-3817/
(From OE-Core rev: 01a1f016a6558566a36098a993adaf4b40e30c78)
Signed-off-by: Chee Yang Lee <chee.yang.lee@intel.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
At second running, there are four new failed case:
d_loaddump f_bigalloc_badinode f_bigalloc_orphan_list f_dup_resize
The test_data.tmp is necessary, but it is deleted by run-ptest.
So it should be restored after testing.
(From OE-Core rev: 496dc586446d09a77923bc191e5305ba06c8f59e)
Signed-off-by: Qiu Tingting <qiutt@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
(cherry picked from commit d0d08dd9a8a179e25b9cfcbac696c1d212a1910c)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Git is a revision control system. Prior to versions 2.30.9, 2.31.8, 2.32.7,
2.33.8, 2.34.8, 2.35.8, 2.36.6, 2.37.7, 2.38.5, 2.39.3, and 2.40.1, by feeding
specially crafted input to `git apply --reject`, a path outside the working
tree can be overwritten with partially controlled contents (corresponding to
the rejected hunk(s) from the given patch). A fix is available in versions
2.30.9, 2.31.8, 2.32.7, 2.33.8, 2.34.8, 2.35.8, 2.36.6, 2.37.7, 2.38.5, 2.39.3,
and 2.40.1. As a workaround, avoid using `git apply` with `--reject` when applying
patches from an untrusted source. Use `git apply --stat` to inspect a patch before
applying; avoid applying one that create a conflict where a link corresponding to
the `*.rej` file exists.
References:
https://nvd.nist.gov/vuln/detail/CVE-2023-25652
Upstream-Status: Backport from 9db05711c9
(From OE-Core rev: 6747482316b8f7839a09bf041d8c11b559f84b44)
Signed-off-by: Hitendra Prajapati <hprajapati@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Git is a revision control system. Prior to versions 2.30.9, 2.31.8, 2.32.7, 2.33.8,
2.34.8, 2.35.8, 2.36.6, 2.37.7, 2.38.5, 2.39.3, and 2.40.1, a specially crafted
`.gitmodules` file with submodule URLs that are longer than 1024 characters can used
to exploit a bug in `config.c::git_config_copy_or_rename_section_in_file()`. This bug
can be used to inject arbitrary configuration into a user's `$GIT_DIR/config` when
attempting to remove the configuration section associated with that submodule. When the
attacker injects configuration values which specify executables to run (such as
`core.pager`, `core.editor`, `core.sshCommand`, etc.) this can lead to a remote code
execution. A fix A fix is available in versions 2.30.9, 2.31.8, 2.32.7, 2.33.8, 2.34.8,
2.35.8, 2.36.6, 2.37.7, 2.38.5, 2.39.3, and 2.40.1. As a workaround, avoid running
`git submodule deinit` on untrusted repositories or without prior inspection of any
submodule sections in `$GIT_DIR/config`.
References:
https://nvd.nist.gov/vuln/detail/CVE-2023-29007
Upstream patches:
528290f8c629198213c9a5bb10fd5ee91cfe60853bb3d6bac5
(From OE-Core rev: db4c152441aebe4c04a7bb7aceb88d8941a6576b)
Signed-off-by: Hitendra Prajapati <hprajapati@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
If a package with a postsints script requires ldconfig, the package class adds
a ldconfig postinst fragment to initialize it before. Systemd has its own
ldconfig.service to initialize it and sometimes if both services are running
at the same time in the first boot, the first one will work, but the second
one will fail with the following error:
ldconfig[141]: /sbin/ldconfig: Renaming of /etc/ld.so.cache~ to /etc/ld.so.cache failed: No such file or directory
This commit adds a ordering dependency between them to make sure that only one
service is running at the same time.
(From OE-Core rev: 1bc254e7969f3d5470bacf9ad9f065d38b7b7fde)
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 4e9d812e127dc6743f52f4881e509e8e2e833afe)
Signed-off-by: Jermain Horsman <jermain.horsman@nedap.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
A parsed MIME header is a map[string][]string. In the common case,
a header contains many one-element []string slices. To avoid
allocating a separate slice for each key, ReadMIMEHeader looks
ahead in the input to predict the number of keys that will be
parsed, and allocates a single []string of that length.
The individual slices are then allocated out of the larger one.
The prediction of the number of header keys was done by counting
newlines in the input buffer, which does not take into account
header continuation lines (where a header key/value spans multiple
lines) or the end of the header block and the start of the body.
This could lead to a substantial amount of overallocation, for
example when the body consists of nothing but a large block of
newlines.
Fix header key count prediction to take into account the end of
the headers (indicated by a blank line) and continuation lines
(starting with whitespace).
Thanks to Jakob Ackermann (@das7pad) for reporting this issue.
Fixes CVE-2023-24534
For #58975Fixes#59267
(From OE-Core rev: daa6aa9c7198a07322f1828a9db457fec86191cf)
Signed-off-by: Vivek Kumbhar <vkumbhar@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Setting a large line or column number using a //line directive can cause
integer overflow even in small source files.
Limit line and column numbers in //line directives to 2^30-1, which
is small enough to avoid int32 overflow on all reasonbly-sized files.
(From OE-Core rev: d1943e6a0ec00653c81cd4c0bb0d6b7e0909094c)
Signed-off-by: Vivek Kumbhar <vkumbhar@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
encoding/xml: replace comments inside directives with a space
Backport from a9cfd55e2b
(From OE-Core rev: 76d855f3d2c250ac85ca6f24bf0e178fb32607f9)
Signed-off-by: Shubham Kulkarni <skulkarni@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
This CVE is specific to Microsoft Windows, ignore it.
Patch fixing it (https://go-review.googlesource.com/c/go/+/446916)
also adds a redundant check to generic os/exec which
could be backported but it should not be necessary as
backport always takes a small risk to break old code.
(From OE-Core rev: 4263f3fda59aacb4f159d2dffb52e5f66249b5e4)
Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
path/filepath: do not Clean("a/../c:/b") into c:\b on Windows
Backport from bdf07c2e16
(From OE-Core rev: 70135bf04eb7173434a7240ddf11639d13aab003)
Signed-off-by: Shubham Kulkarni <skulkarni@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
The patch for CVE-2021-3929 applied on dunfell returns a value for a
void function. This results in the following compiler warning/error:
hw/block/nvme.c:77:6: error: void function
'nvme_addr_read' should not return a value [-Wreturn-type]
return NVME_DATA_TRAS_ERROR;
^ ~~~~~~~~~~~~~~~~~~~~
In newer versions of qemu, the functions is changed to have a return
value, but that is not present in the version of qemu used in “dunfell”.
Backport some of the patches to correct this.
(From OE-Core rev: 4ad98f0b27615ad59ae61110657cf69004c61ef4)
Signed-off-by: Gaurav Gupta <gauragup@cisco.com>
Signed-off-by: Gaurav Gupta <gauragup@cisco.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Disable cmd/internal/moddeps test, since this update includes PRIVATE
track fixes.
Backport from 5c3e11bd0b
(From OE-Core rev: 53a303fb5908edaf29e35abb08fff93e7c0ff92c)
Signed-off-by: Shubham Kulkarni <skulkarni@mvista.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
import patches from ubuntu to fix
CVE-2020-15469
CVE-2020-15859
CVE-2020-17380
CVE-2020-35504
CVE-2020-35505
CVE-2021-3409
CVE-2022-26354
https://git.launchpad.net/ubuntu/+source/qemu/tree/debian/patches?h=ubuntu/focal-security
Combine patches for both CVE-2020-25085 and CVE-2021-3409 also fix CVE-2020-17380.
so mark CVE-2020-17380 fixed by CVE-2021-3409 patches. CVE-2020-17380 patch backported since
oecore rev 6b4c58a31ec11e557d40c31f2532985dd53e61eb.
(From OE-Core rev: 3ee2e9027d57dd5ae9f8795436c1acd18a9f1e24)
Signed-off-by: Chee Yang Lee <chee.yang.lee@intel.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Backport of gcc upstream commit 2b2d97fc545635a0f6aa9c9ee3b017394bc494bf
to gcc release 9.5.0. This fix is available to all release-branches
except releases/gcc-9 because upstream do not support gcc-9 now.
(From OE-Core rev: 9779b66162a014f26594bdde43afdc4332617240)
Signed-off-by: Virendra Thakur <virendrak@kpit.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This test assumes that if a child process writes one line to stderr and
then another line to stdout, and stderr is redirected to stdout, that
the order the lines will be read is stable.
This isn't the case and occasionally the lines will be read in a
different order. Change the test to ignore line ordering.
[ YOCTO #14469 ]
(From OE-Core rev: 2c9fe8c3bb1cc1883c7bd445d019b2107e85ab2b)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 1ddbe4d2bd8d8da10dac8a054f130fcd1d242219)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a PACKAGECONFIG option for slirp, defaulting to internal. This avoids
the presence of libslirp on the host causing qemu to link against that
instead breaking reproducibility and usability of the binary on hosts
where the library isn't present.
We need to add it to PACKAGECONFIG by default since users do expect slirp
to be enabled in the wider community.
Note: qemu version 4.2.0 doesn't support an "internal" option for
enable-slirp, so use "git" instead which uses the same configure
code path, avoids host libslirp contamination and forces use of the
qemu internal slirp implementation.
(From OE-Core rev: e5dc03e4a3b71ff144896a8ce56a34b8677e8e27)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 5a9a64132bf5ecac9d611d29751226a466c4a2c1)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
