The info->resolvetls might be NULL according to src/prelink.c:
[snip]
/* Dynamic linker does not depend on any other library,
all symbols resolve to themselves with the exception
of SHN_UNDEF symbols which resolve to 0. */
if (info->symtab[r_sym].st_shndx == SHN_UNDEF)
{
info->resolveent = NULL;
info->resolvetls = NULL;
[snip]
So we must check it before use its members, otherwise, there might be
Segmentation fault error.
Fixed:
MACHINE = "qemumips"
IMAGE_INSTALL_append = " qemu"
$ bitbake core-image-minimal
[snip]
| /path/to/qemumips-poky-linux/core-image-minimal/1.0-r0/temp/run.prelink_image.1000: line 111: 1010 Segmentation fault (core dumped)
[snip]
(From OE-Core rev: 7f5d6f6b70cb1eb486d3ff5c029fc3c6202fc826)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This reverts commit bd8470dcaebddf37bf889d3dffd3dcb53d367f23.
It breaks prelink since its using the master branch which doesn't contain
the cross changes.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The new version of prelink includes the fix for a segmentation fault due to
using e_shnum vs e_shstrndx.
In addition the following changes are incorporated:
2018-08-29 Khem Raj <raj.khem@gmail.com>
* src/elf.h: Add RISC-V defines
* src/rtld/dl-tls.c: Add RISC-V support
* src/rtld/rtld.c: Add RISC-V support
2018-08-29 Robert Yang <liezhi.yang@windriver.com>
* src/dso.c: use ehdr.e_shstrndx as index
2017-06-20 Andrew Stubbs <ams@codesourcery.com>
* src/arch-x86_64.c (x86_64_prelink_conflict_rela): Also convert
R_X86_64_32 conflicts to R_X86_64_IRELATIVE for ifuncs.
* src/arch-x86_64.c (PL_ARCH(x32)): Set mmap_end to 0x60000000.
2017-06-20 Joseph Myers <joseph@codesourcery.com>
* src/rtld/rtld.c (load_ld_so_conf): Add argument use_x32.
(main): Update call to load_ld_so_conf.
2017-06-20 Kyle Russell <bkylerussell@gmail.com>
* src/rtld/rtld.c: Add missing DT_NEEDED DSOs to needed_list
(From OE-Core rev: 1f0568c6b8982f814f630193fb5114db489afcaf)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #12791]
According to struct elf32_hd, the e_shnum is section header number, and the
index is e_shstrndx, not e_shnum.
This can fix segmention fault when handle libqb.so.0.18.2 from libqb_1.0.3.
It fails to handle libqb.so.0.18.2 and get errors:
Symbol section index outside of section numbers
Then segmentation fault, this is because the e_shnum is 34, while e_shstrndx is
27 (it would be 33 when no errors), I've checked several elf files to confirm
that the ones after e_shstrndx is NULL, so use e_shstrndx should be correct.
Fixed:
MACHINE="qemux86-64"
IMAGE_INSTALL_append = " libqb" #libqp is from meta-openembedded
$ bitbake core-image-minimal
Segmention fault
(From OE-Core rev: b7c291ee6532cba845ee6bfbbaa21076a2b2cbe5)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Update prelink to the newest commit on the cross_prelink ref. This
includes the following changes:
05aeafd053 Disable automatic generation of prelink.pdf
aa2985eefa src/rtld: Add MicroBlaze support based on glibc-2.24
62f80843f8 README: update information on reloc8/reloc9 failures
The primary purpose of this update is to enable the added MicroBlaze
rtld support so as to enable gobject-introspection for MicroBlaze.
(From OE-Core rev: 8cab8f79b73e72256a2d7940f4973ebfd36e232b)
Signed-off-by: Nathan Rossi <nathan@nathanrossi.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This greatly reduces the amount of recipes for which upstream
version check fails: from about 30 to about 8.
(From OE-Core rev: bb9066670a6096aa4134bebc84132b4eb3964b6d)
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>
The behavior before this change was to check USER_CLASSES and adjust
the install script to return either exit 0 (don't do anything) or
exit 1 (run on first boot). This enabled a user to include the prelink
package without enablign the image-prelink bbclass and get a first boot
prelink.
Checking USER_CLASSES is not desired, as an image should be able to simply
inherit the image-prelink and get the same type of behavior. Modifying
the recipe based on the inclusion of a class is a bad idea as it makes
this style work more difficult. So we move to a more defined strategy
based on exist uses. (That we know of...)
If we ae doing a cross install, we want to avoid prelinking.
Prelinking during a cross install should be handled by the image-prelink
bbclass. If the user desires this to run on the target at first boot
they will need to create a custom boot script.
[YOCTO #11169]
(From OE-Core rev: e31c9d32072b9cf62c0e9e55b4d421849d3d489b)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
getVar() now defaults to expanding by default, thus remove the True
option from getVar() calls with a regex search and replace.
Search made with the following regex: getVar ?\(( ?[^,()]*), True\)
(From OE-Core rev: 7c552996597faaee2fbee185b250c0ee30ea3b5f)
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>
Currently, prelink doesn't work unless base_libdir/libdir match
its hardcoded values. This patch manipulates those paths so that
they match the values set in the variables and handles multilib
configurations too. The manipulations only happen in the target
case, if needed.
[YOCTO #10282]
(From OE-Core rev: 6779990021530129a78eb73db122e3976f687c7e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fix the following issues:
IA32 / ARM - Resync to glibc-2.22, fix a mismatch w/ glibc's ld.so
MIPS - Ignore the new SHT_MIPS_ABIFLAGS
ARM - Fix missing ARM IFUNC support chunk
Also upstream prelink project no longer has a 'trunk' directory.
(From OE-Core rev: c725328f2ab5c9b220c552ed37c0d24b098a218d)
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>
[YOCTO #7056]
Adds support for armhf dynamic linker
Update prelink-rtld to glibc 2.20 base
Fix testsuite to work with latest autotools
(From OE-Core rev: c9e4867a605a8007f569ed6ec4cf445111a533c1)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Update to the current version of cross-prelink development:
faa069deec99bf61418d0bab831c83d7c1b797ca
This adds a number of minor features, such as S390 support (irrelevant)
to ARM TLS DESC relocations.
(From OE-Core rev: e87a56bbdfbae56bc027ab66b350cc8fa7b65810)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The base_contains is kept as a compatibility method and we ought to
not use it in OE-Core so we can remove it from base metadata in
future.
(From OE-Core rev: d83b16dbf0862be387f84228710cb165c6d2b03b)
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The revision specified is on the cross_prelink branch so mark the
url accordingly else the fetch can fail in the no network case.
(From OE-Core rev: 109166f1c0a84bcbbcd9d6adff7c16ddee08b08b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It's not necessary to specify the protocol parameter when it's the
default protocol for the fetcher, e.g. the default protocol for
git fetcher it git, "protocol=git" isn't needed.
(From OE-Core rev: a2bab241c64428d5109c3c5ac5de4463fbad70c5)
Signed-off-by: Jackie Huang <jackie.huang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Sync to latest cross-prelink, as of Aug 1, 2013.
Drop the PR number, no longer needed.
(From OE-Core rev: b59d256349cc0ac19357158be8e63bd52ab9fb51)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Compatible with automake-1.13.
(From OE-Core rev: 3d8f35a121d54b7a1b825b266a553752e22bf708)
Signed-off-by: Marko Lindqvist <cazfi74@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The latest fixes an obscure problem with prelinking on PowerPC, as well
as general updated.
(From OE-Core rev: 7759c9a2c5dfdc1cae5540a5985632c78abbac89)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The latest version of prelink corrects a problem when attempting
to prelink MIPS architectures.
2012-01-04 Mark Hatle <mark.hatle@windriver.com>
* exec.c: Check that a section is larger then 0 bytes when
determining the section to segment mapping. This matches
the behavior of elfutils - readelf. Otherwise an empty
PROGBITS section at the end of a segment will cause a
failure.
[YOCTO #1463]
(From OE-Core rev: 09a70c55e590d169b8a3b4b89853c96b7b977fc0)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Prelinking for x32 image showed that it was ignoring libraries
located at locations like /usr/libx32. Like that mips n32 has
library locations set as lib32
This commit modifies prelink.conf to look at libraries also
located at libx32 & lib32 locations.
Thanks to Mark Hatle for suggesting the fix.
(From OE-Core rev: c02b9de25b405c81da0f7bebd07423e8cee14eb7)
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
New version includes:
2011-12-08 Mark Hatle <mark.hatle@windriver.com>
* rtld/rtld.c: Fix an issue where missing objects would trigger
an assert in dl-version.c
* rtld/rtld.h: Add _dl_new_object prototype
* rtld/rtld.c: Add support for $ORIGIN, $PLATFORM and $LIB.
Note: $PLATFORM = ""
These fixes are needed to prevent prelink-rtld from crashing when
running the new unsafe-references-in-binaries QA test.
(From OE-Core rev: 1a80d91b163869c9cb1d12d502aa8bc25e8782fa)
Signed-off-by: Scott Garman <scott.a.garman@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #1473]
From the upstream cross prelink respository:
Fix printf typecast problem
Without this fix the printed TLS(...) information is incorrect, which
causes the core prelinker code to incorrectly prelink. This bug triggers
primarily on x86 (32-bit) hosts, targeting x86 (32-bit) targets and arm
targets.
(From OE-Core rev: 2fc778068ebbd97b0fb473fb35a3eca03da86542)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
[YOCTO #1374] - Fixes underlying issue
[YOCTO #1331] - Detects bad GNU_HASH tables
Update to the latest upstream version. This includes a number of
generic fixes, as well as a complete refresh of the prelink-rtld.
(Prelink-rtld is similar to the ld.so/ldd functionality on a normal
system.)
The new prelink-rtld update syncs to eglibc-2.13 and enables new
symbol resolution, include GNU_UNIQUE, that was causing problems on
x86_64 and potentially other architectures with C++.
An assert has been added to also detect a bad GNU_HASH situation
which was detected on some PPC64 systems. The underlying problem is
within the toolchain, however prelink was making it worse leaving the
problem undetected.
(From OE-Core rev: 4f401134694bbc1de532aaaab6c9a761649725a1)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If image-prelink is being used, the system will automatically prelink
the target image. This avoids the need to run the postinst prelink
script at first boot. However, if the user has not enabled image
prelinking -- then we do enable the script to run on first boot.
(From OE-Core rev: a0ba05596a704a006ee7b97392c20b871085b1e6)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The latest cross-prelink version tracks the prelink r190 SVN.
Fix two minor issues as well. If we remove the package, we would
unprelink the filesystem after the rm was finished. This is
incorrect.. we need to do this prerm.
Move the cron and cron configuration file to a new package split.
This item is not valid for most embedded users.
(From OE-Core rev: e8da02aeae4b3c1c2a805d63d6d0919d3034aba6)
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
in this case, those non poky distro can also use these recipe normally
(From OE-Core rev: 0a57bd226cdb8332707fa0f46fcf0b067f03701a)
Signed-off-by: Yu Ke <ke.yu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
RPM has the ability to validate files that have been prelinked, however
the necessary configuration and staging was not done properly. Resolve
this issue by fixing the macro paths, providing the missing RPM macro,
and correcting a defect in the way the prelink image class was working
with the necessary configuration file.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
This issue is really do to the timestamps of files being different
enough and that the make of the doc directory is being triggered by
this.
This is partly due to the new git fetcher laying the files down
differently. Once this issue is resolve, this can be reverted.
Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Having one monolithic packages directory makes it hard to find things
and is generally overwhelming. This commit splits it into several
logical sections roughly based on function, recipes.txt gives more
information about the classifications used.
The opportunity is also used to switch from "packages" to "recipes"
as used in OpenEmbedded as the term "packages" can be confusing to
people and has many different meanings.
Not all recipes have been classified yet, this is just a first pass
at separating things out. Some packages are moved to meta-extras as
they're no longer actively used or maintained.
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
