Richard managed to reproduce a pseudo bug that pciutils in a minimal
test case[1] which confirmed that it was the symlink creation causing
issues.
The pciutils Makefile has several installation targets:
- install: binaries, manpages, shared library
- install-lib: headers, library symlinks, shared library
We need to run both targets to install the full set of files we want to
be installed, but notably they both create the .so symlink and as they're
running in parallel this is what triggers the bug in pseudo.
Until the bug has been resolved (or the Makefiles don't duplicate rules),
just run the two targets separately.
[ YOCTO #14957 ]
[1] https://lore.kernel.org/openembedded-core/20250319133457.806384-1-richard.purdie@linuxfoundation.org/T/#u
(From OE-Core rev: a5fc49fafe910a25be7372c82bfbd7876871ce3f)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The current pciutils does not have this textrel problem any more.
(From OE-Core rev: 48fd3134097ce81c94c1ce67f218d83f1a5c2f6f)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog:
===========
* lspci decodes CXL 1.1 device link status information.
This requires a recent kernel which exports rcd_* atributes via
sysfs.
* Further development of the pcilmr (the link margining utility)
* Dump parsing supports 6-digit domain numbers.
* Bug fixes in PCIe link state reporting.
* Decode more fields in PCIe AER capability.
* Fixed build on Linux systems with musl libc.
* Updated pci.ids.
(From OE-Core rev: cdd6e828cae9b3b1fcf866dfdcf6409426abfe2c)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The upstream Makefiles are bespoke so need to be handheld, but this
recipe has workarounds for code removed a decade ago.
Add more PACKAGECONFIGs for the other configuration options (zlib, kmod),
and use hwdb by default (non-systemd builds also have hwdb, as it's part
of eudev too).
Instead of using 'make -e', pass the variables that we want the Makefile
to respect explicitly (CC, CFLAGS, etc).
Construct an appropriate HOST so that lib/configure does the right thing
without needing patches.
This now results in text relocations (apparently from the versioned
symbols in libpci), so add an INSANE_SKIP for that.
Update the homepage URL.
(From OE-Core rev: e943b74ffe0fd2c37e35540a2d6bd92a1dd61a3a)
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>
configure.patch is rebased against the new version.
(From OE-Core rev: 88ab0efeda625892f393576fbcc87e9b517103f5)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
configure.patch
refreshed for 3.11.1
Changelog:
===========
* Fixed wrong API version in lib/pci.h.
* Updated README.Windows.
* Fix compilation on Windows.
* update-pciids now supports XZ compression. If libpci is configured
with support for compression, all downloaded files are recompressed
as gzip. Otherwise they are stored as plain text.
* update-pciids now sends itself as the User-Agent.
* Added a pcilmr utility for PCIe lane margining.
* Re-factored access to i386 ports on all relevant platforms.
* Added i386 port access on OpenBSD.
* Back-ends for Windows received many bug fixes and improvements.
* ECAM back-end now scans ACPI and BIOS memory faster.
* Linux systems without pread/pwrite are no longer supported
as they are hopefully long gone. This helps avoid the tricky check
for presence of pread which was found to fail on musl libc.
* Improved decoding of PCIe control and status registers.
* Decoding of CXL capabilities now supports up to CXL 3.0.
* lspci now displays interrupt message numbers consistently across
different capabilities.
* Cache of IDs resolved via DNS, which was located in ~/.pci-ids
by default, is now stored according to the XDG base directory
specification in $XDG_CACHE_HOME/pci-ids.
* All source files now have SPDX license identifiers.
* Internal: The "aux" fields of structs pci_access and pci_dev
reserved for use by back-ends were renamed to backend_data to better
reflect their meaning.
* As usually, various minor bug fixes and updated pci.ids.
(From OE-Core rev: bbda388857284612190b58abe8d73c93e39d1574)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changelog (see https://github.com/pciutils/pciutils/blob/master/ChangeLog):
* Fixed bug in definition of versioned symbol aliases in shared libpci,
which made compiling with link-time optimization fail.
* Filters now accept "0x..." syntax for backward compatibility.
* Windows: The cfgmgr32 back-end which provides the list of devices
can be combined with another back-end which provides access to
configuration space.
* ECAM (Enhanced Configuration Access Mechanism), which is defined by
the PCIe standard, is now supported. It requires root privileges,
access to physical memory, and also manual configuration on some
systems.
* lspci: Tree view now works on multi-domain systems. It now respects
filters properly.
* Last but not least, pci.ids were updated to the current snapshot of
the database. This includes overall cleanup of entries with non-ASCII
characters in their names -- such characters are allowed, but only if
they convey interesting information (e.g., umlauts in German company
names, but not the "registered trade mark" sign).
(From OE-Core rev: 04b5aa6c73bc3b9f0be2b28acfa0d66c2ecca310)
Signed-off-by: Trevor Gamblin <tgamblin@baylibre.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
configure.patch
refreshed for 3.9.0
Changelog:
==========
* We decode Compute Express Link (CXL) capabilities.
* The tree mode of lspci is now compatible with filtering options.
* When setpci is used with a named register, it checks whether
the register is present in the particular header type.
* Linux: The intel-conf[12] back-ends prefer to use ioperm() instead
of iopl() to gain access to I/O ports.
* Windows: We have two new back-ends
One uses the NT SysDbg interface, the other uses kldbgdrv.sys
(which is a part of the Microsoft WinDbg tool).
* Windows: We support building libpci as a DLL. Also, Windows
binaries now include meta-data with version.
* Hurd: The Hurd back-end works again.
* mmio-conf1(-ext): Added a new back-end implementing the intel-conf1
interface over MMIO. This is useful on some ARM machines, but it
requires manual configuration of the MMIO addresses.
* As usually, updated pci.ids to the current snapshot of the database.
(From OE-Core rev: 8a784f5aa7ceec1866ef12f2bedf4e8b4e4f269f)
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Rebase configure.patch
Drop symlink creation; make install already does it.
(From OE-Core rev: 8b35bad89eafd51d49438725e781450d7907d0f2)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
An automated conversion using scripts/contrib/convert-spdx-licenses.py to
convert to use the standard SPDX license identifiers. Two recipes in meta-selftest
were not converted as they're that way specifically for testing. A change in
linux-firmware was also skipped and may need a more manual tweak.
(From OE-Core rev: ceda3238cdbf1beb216ae9ddb242470d5dfc25e0)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is the result of automated script conversion:
scripts/contrib/convert-overrides.py <oe-core directory>
converting the metadata to use ":" as the override character instead of "_".
(From OE-Core rev: 42344347be29f0997cc2f7636d9603b1fe1875ae)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
refresh the following patch:
configure.patch
(From OE-Core rev: 498e0f83040fe2e83805d8b13d69c17c74a520ba)
Signed-off-by: Wang Mingyu <wangmy@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Move the recipe specific over-ride for ASNEEDED into the recipe to
make it more apparent that the over-ride is being applied (and that
it should be re-checked on version updates, etc).
(From OE-Core rev: b7b63b2681a1de0ecb0e09612913370cb9934d38)
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The patch tool will apply patches by default with "fuzz", which is where if the
hunk context isn't present but what is there is close enough, it will force the
patch in.
Whilst this is useful when there's just whitespace changes, when applied to
source it is possible for a patch applied with fuzz to produce broken code which
still compiles (see #10450). This is obviously bad.
We'd like to eventually have do_patch() rejecting any fuzz on these grounds. For
that to be realistic the existing patches with fuzz need to be rebased and
reviewed.
(From OE-Core rev: 7d56245a4859727f85dc5ae6f881d0783dee1ce1)
Signed-off-by: Ross Burton <ross.burton@intel.com>
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>
Patches patching content from other patches is bad. This folds the three
patches into one, improves some of what its doing to be a little
cleaner and adds a proper description to the patch header.
This also moves the STRIP override to the configure commandline rather
than patching and handles passing libdir the same way.
(From OE-Core rev: 186b30798cf74b738483e8708157cac628c1d3cd)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some of these are clearly dead, e.g. one binutils patch reverts the effects
of the earlier one.
This also removes the uclibc site files. We now have mechanisms to allow these
to be extended from another layer should someone ever wish to do that.
(From OE-Core rev: e01e7c543a559c8926d72159b5cd55db0c661434)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The PACKAGECONFIG's value doens't go into EXTRA_OECONF, but
PACKAGECONFIG_CONFARGS.
Fixed:
pciutils-3.5.1: libpci rdepends on libudev, but it isn't a build dependency, missing eudev in DEPENDS or PACKAGECONFIG? [build-deps]
pciutils-3.5.1: pciutils rdepends on libudev, but it isn't a build dependency, missing eudev in DEPENDS or PACKAGECONFIG? [build-deps]
(From OE-Core rev: d941d66d714545eae589115db48f1243399711f2)
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>
The openssl recipe currently relies on EXTRA_OEMAKE having been set to
"-e MAKEFLAGS=" in bitbake.conf to operate. It is necessary to make this
explicit so that the default in bitbake.conf can be changed.
(From OE-Core rev: 92850650d32fcf609e7fa7a291774fc2bbe5ef5b)
Signed-off-by: Mike Crowe <mac@mcrowe.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Default PACKAGECONFIG value to "hwdb" if 'systemd' is in DISTRO_FEATURES.
(From OE-Core rev: 9a7d1149c5f97c8f7b8fef12b4f77ae7c6d88b2c)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add PACKAGECONFIG for 'hwdb' to fix the following QA warning.
WARNING: QA Issue: libpci rdepends on libudev, but it isn't a build dependency? [build-deps]
WARNING: QA Issue: pciutils rdepends on libudev, but it isn't a build dependency? [build-deps]
(From OE-Core rev: 9188a2672013cb6960590f03f753178999fe4c3c)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
lib-build-fix.patch is removed as it doesn't seem useful.
do_compile_prepend is removed because it's not useful for the new version.
The following line is removed as I don't see any problem without it.
PARALLEL_MAKE = ""
(From OE-Core rev: 0df811d5afdb947bfef98f7c53268259c99b88be)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* Move package-specific defitions to the end
* Define pciutils-ids package as ${PN}-ids
* Add SUMMARY and tweak DESCRIPTION for pciutils-ids package
(From OE-Core rev: c8c39c78878cdffe47b376b6211dd1ae45fd59d2)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Remove all PR = "r0" from all .bb files in oe-core. This was done
with the command sed -e '/^PR.*=.*r0\"/d' recipes*/*/*.bb -i
We've switching to the PR server, PR bumps are no longer needed and
this saves people either accidentally bumping them or forgetting to
remove the lines (r0 is the default anyway).
(From OE-Core rev: 58ae94f1b06d0e6234413dbf9869bde85f154c85)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
pcimodules can not work due to no modules.pcimap file, and has been
replaced by "lspci -k", so we can remove it.
Update lib-build-fix.patch since remove pcimodules-pciutils.diff.
[YOCTO# 5210]
(From OE-Core rev: 57bec61b05c0fea42f988dd52942c7d5ce8833fd)
Signed-off-by: Roy Li <rongqing.li@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
configure.patch was missing a closing quote in a sed expression which
causes script execution to error on less forgiving shells such as dash.
(From OE-Core rev: ad72484a2783afa948966263629006558fbde476)
Signed-off-by: Joshua Lock <josh@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In mirrors.bbclass we point into /pub but we don't in bitbake.conf.
All uses of KERNELORG_MIRROR look into /pub anyhow, so lets make
use of that.
(From OE-Core rev: 67a0c8f48b5ef2ae5fc712c9204e4e99818c8134)
Signed-off-by: Tom Rini <tom_rini@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
