| /[gentoo-x86]/eclass/linux-mod.eclass |
Parent Directory
| 
Revision Log
| Links to HEAD: | (view) (download) (annotate) |
Remove use of sed in linux-mod.eclass. Replace with bash.
Always ensure MODULES_OPTIONAL_USE is in IUSE.
MODULES_OPTIONAL_USE makes it possible to optionally use linux-mod without introducing dependencies on virtual/linux-sources or virtual/modutils where unwanted.
Punt unused if and else wrt bug 462926, Comment #16 and up
Control pulling in linux-mod dependencies based on VIRTUAL_MODUTILS variable, set in eg. sys-apps/kmod
Stop pulling in virtual/linux-sources when used for sys-apps/kmod by angry_vincent
Don't pull modutils if using the eclass from sys-apps/kmod
Remove support for module-init-tools and Linux 2.4. Only works with kmod and 2.6 now.
Remove redundant DESCRIPTION variable settings.
Minor typo fix to documentation line. Bug #419469
Remove deprecated and unrecommended parameter -r from depmod
Bug #404193: Depend on virtual/modutils instead of sys-apps/module-init-tools.
fix random bugs in eclass documentation, and convert to new @AUTHOR tag
drop old comment now that eutils.eclass no longer contains these funcs
Update linux-mod.eclass for building modules on Linux 3.0* kernels.
for EAPI>=4, rely on PMS MERGE_TYPE rather than portage non-spec EMERGE_FROM; for EAPI<4, use EMERGE_FROM since it's acceptable there
Replace dependency on virtual/modutils, bug 358891. Whitespace.
Bug #312075: The ~ was getting expanded to the optional stuff got duplicated and that sucked, giving a bad warning sometimes.
Better version of the logic for bug #286292.
Bug #286292: Try to make kernel module binpkgs NOT need kernel sources by using a different path in pkg_setup.
Bug #258209: Fix quoting of directories.
Bug #283320: The dependancy for linux-sources now moves into linux-mod, where it is actually required.
Name module configuration files properly as per bug #277835. Thanks to MATSUU Takuto for the patch
share Module.symvers between all compiled modules, patch from Peter Volkov in bug #237618
Remove RESTRICT=distcc for now, its obvious it will never be an option when a 3rd party decides what goes in our PMS.
documentation tweaks from Fabio Rossi
Add bug #120001 to comment for RESTRICT=distcc
Forgot to add a comment to the eclass for commit: Add distcc to RESTRICT in linux-mod.eclass bug #167844
Add distcc to RESTRICT in linux-mod.eclass bug #167844
allow for relative srcdir directory, thanks to Alon Bar-Lev in bug #198642
add some fantastic documentation from Fabio Rossi in bug #217560
add some documentation, patch from Eric Brown in bug #190934
use set_arch_to_kernel/set_arch_to_portage rather than managing ARCH ourselves #194380
pass CROSS_COMPILE=${CHOST}- to build process #194380
allow KERNEL_CC env var to override get-KERNEL_CC()
fix if that was missing then
Install modprobe.d instead of modules.d files for kernel-2.6, thanks Cardoe for helping
Correctly handle variables with spaces being over and under interpreted when the emake is called. Symptoms of this are weird targets being used for make.
Don't require a configured kernel to detect kernel version, or if all CONFIG_CHECK items are prefixed with ~
Replace references to deprecated ${IMAGE} with equivalent ${D}.
These seem to be the last ${IMAGE} references in the whole tree.
Change to ${S} before starting the module build (see #192549 with thanks to Elias Probst <mail at eliasprobst.eu>).
kernel related eclasses are maintained by kernel-misc@gentoo.org.
modules-update -> update-modules
Use update-modules thanks to jakub in bug 174745
tweak grepping of moduledb to avoid errors when moduledb hasnt actually been installed
Fix bad sed arguments thanks to Georgi Georgiev in bug #165230
Apply whitespace checks from repoman to eclasses.
shorten compile part a bit and add debug-print-funtion calls thanks to Petteri Räty <betelgeuse@gentoo.org> in bug 156186
Add me as maintainer and commit kernel/userland ABI patch from bug 133382 thanks to robbat2
remove pcmcia specific functons, bug 140116, also solves bug 122868
kill off pointless whitespace
prefix linux-only DEPENDs with kernel_linux? and only run needed functions in pkg_postinst, helps to fix bug 133548
Allow for cross-compile/distcc distribution of modules.
Fixes bug #12645 and misc.
set ABI= in src_compile; bug 119012
We need to inherit eutils before linux-info so that set_arch_to_kernel and set_arch_to_portage functions from linux-info are used instead of the ones in eutils. Fixes bug #127506.
Fixes QA bug #124324, thanks Ciaran
Fixing a stupid little bug.
Fixed bug #120233.
Fixed bug #122101.
add some basic sanity checks
*** empty log message ***
Commented out check_vermagic with permission from johnm until a fixed version for arches with different userland and kernel compilers is available.
Changing some check_vermagic logic to support KV_DIR better. Fixing MODULESD error reported by genstef
Adding in vermagic check. This is a little crude, but picks up on the most common problem
Fixes bug #106092
Remove display_postinst() function as it was causing more trouble than good. Most modules will be auto-loaded by coldplug/hotplug.
Fix DEPEND.blacklist: virtual/modutils for linux-mod.eclass ebuilds.
QA: Apply whitespace checks from repoman to eclasses.
add a note regarding IUSE and pcmcia
Removed IUSE=pcmcia - please contact the kernel herd before modifying.
Fixed DEPEND/RDEPEND issue.
add IUSE=pcmcia
Remove ECLASS= and INHERITED= which haven't been necessary for a long time now
Remove ECLASS= and INHERITED= which haven't been necessary for a long time now
minor change
oops, missed to commit thjis earlier. Handle moduledb a little tidier
changing moduledb dir to correspond with FHS and closes bug #97796
sys-apps/pcmcia-cs => virtual/pcmcia.
Somehow the IUSE=pcmcia line sneaked in again...
Additions to compliment module-rebuild. Fixes bug #34921, thanks go to Paul Varner for inspiration
Reverting pcmcia iuse change
Adding pcmcia flag. Closes #90025, Cheers Georgi
No virtual yet, depend on directly on sys-apps/pcmcia-cs.
merging pcmcia support into linux-mod, now awaiting pcmcia-cs changes and were good to go
Fixing small bug (#86650) with linux-mod/convert_to_m, thanks to Stefan Schweizer
linux-mod.eclass: fixed grep in update_modules() so it's not a subshell, and doesn't output when no files are found
Changing array substitutions to _ not -, this prevents the cosmetic bad substitution errors
gagging grep
Removing spurious einfo
Fixing linux-mod modules.d generation
Multiple fixes.
Changing detection of module.d/* file to see if the file is actually setting anything. if it isnt, dont bother running modules-update
checking for modules.d entry before we run modules-update
unset libdir srcdir objdir since they may contain settings for previous modules.
few minor fixes. added support for econf if needed
Changing ARCH unset in src-compile
Update to be less specific with @OPTION:module
Added support for CONFIG_CHECK=@OPTION:module which removes modules from MODULE_NAMES if the support is already in the kernel
adding support for objdir to MODULE_NAMES
changing back to virtual. if this causes problems, then please look at changing the default virtual for that profile to module-init-tools
Changing build logic. now we dont build BUILD_TARGETS more than once, in the same directory. will still support MODULENAME_TARGETS when this logic is available
Fixed bugs: 75034 76849 74655
exporting src_install
Adding support for kernel_is operators, and removing on behalf of line from modules.d regeneration.
adding extra documentation, and the ability to over-ride.
adding support for automated modules.d file generation
adding documentation, and changing the MODULE_NAMES parsing slightly
Ok, last commit tonight, I promise! removing quotes from MODULE_NAMES loops.
moving make clean and fixing SUBDIR typos
Adding support for BUILD_TARGETS
Moving get_version to linux-info pkg_setup and calling from linux-mod pkg_setup when needed.
Fixing up support for KBUILD_OUTPUT.
add support for CHECK_CONFIG=
module-updates addition
Fixing permissions
Updating kernel-2 to properly honour strict patch levels with *.diff , adding linux-* in preparation for the 'Big Module Move of 2004'
This form allows you to request diffs between any two revisions of this file. For each of the two "sides" of the diff, select a symbolic revision name using the selection box, or choose 'Use Text Field' and enter a numeric revision.
| ViewVC Help | |
| Powered by ViewVC 1.1.20 |