d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
b5c2be723b
freebsd-nq/sys/conf/kmod.mk

486 lines
13 KiB
Makefile
Raw Normal View History

A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
# From: @(#)bsd.prog.mk 5.26 (Berkeley) 6/25/91
$Id$ -> $FreeBSD$
1999-08-28 00:22:10 +00:00
# $FreeBSD$
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# The include file <bsd.kmod.mk> handles building and installing loadable
# kernel modules.
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
#
# +++ variables +++
#
# CLEANFILES Additional files to remove for the clean and cleandir targets.
#
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# EXPORT_SYMS A list of symbols that should be exported from the module,
# or the name of a file containing a list of symbols, or YES
# to export all symbols. If not defined, no symbols are
# exported.
#
# KMOD The name of the kernel module to build.
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
The kernel is now known as `kernel.ko' and it and its matching modules live in ``/boot/kernel/''.
2000-09-05 22:37:46 +00:00
# KMODDIR Base path for kernel modules (see kld(4)). [/boot/kernel]
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# KMODOWN Module file owner. [${BINOWN}]
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# KMODGRP Module file group. [${BINGRP}]
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# KMODMODE Module file mode. [${BINMODE}]
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
1) s/mod(un)?load/kld$1load/ 2) s/MODLOAD/KMODLOAD/ to be consistent with the rest of the variables (KMOD, KMODOWN, KMODGRP, etc) and definition of MODLOAD/UNLOAD in the Makefile of the ATAPI module 3) textual fixups
1999-09-06 20:11:59 +00:00
# KMODLOAD Command to load a kernel module [/sbin/kldload]
add variables MODLOAD and MODUNLOAD for modlad/modunload commands
1996-10-06 22:10:35 +00:00
#
1) s/mod(un)?load/kld$1load/ 2) s/MODLOAD/KMODLOAD/ to be consistent with the rest of the variables (KMOD, KMODOWN, KMODGRP, etc) and definition of MODLOAD/UNLOAD in the Makefile of the ATAPI module 3) textual fixups
1999-09-06 20:11:59 +00:00
# KMODUNLOAD Command to unload a kernel module [/sbin/kldunload]
add variables MODLOAD and MODUNLOAD for modlad/modunload commands
1996-10-06 22:10:35 +00:00
#
In the unload target, check that the module is loaded first. Add a reload target which unloads and then loads the module.
2016-01-01 03:59:09 +00:00
# KMODISLOADED Command to check whether a kernel module is
# loaded [/sbin/kldstat -q -n]
#
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# PROG The name of the kernel module to build.
# If not supplied, ${KMOD}.ko is used.
#
# SRCS List of source files.
#
firmware(9) is a subsystem to load binary data into the kernel via a specially crafted module. There are several handrolled sollutions to this problem in the tree already which will be replaced with this. They include iwi(4), ipw(4), ispfw(4) and digi(4). No objection from: arch MFC after: 2 weeks X-MFC after: some drivers have been converted
2006-01-29 02:52:42 +00:00
# FIRMWS List of firmware images in format filename:shortname:version
#
Fix typo in comment. Reported by: thompsa
2007-03-04 19:35:59 +00:00
# FIRMWARE_LICENSE
- Add Intel firmwares for Intel PRO/Wireless LAN 2100/2200/2915 cards in a uuencoded format along with their respective LICENSE files. - Add new share/doc/legal directory to BSD.usr.dist mtree file. This is the place we install LICENSE files for restricted firmwares. - Teach firmware(9) and kmod.mk about licensed firmwares. Restricted firmwares won't load properly unless legal.<name>.license_ack is set to 1, either via kenv(1) or /boot/loader.conf. Reviewed by: mlaier, sam Permitted by: Intel (via Andrew Wilson) MFC after: 1 month
2007-03-02 11:42:56 +00:00
# Set to the name of the license the user has to agree on in
# order to use this firmware. See /usr/share/doc/legal
#
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# DESTDIR The tree where the module gets installed. [not set]
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
#
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
# +++ targets +++
#
# install:
Drop <bsd.man.mk> support from <bsd.kmod.mk>. Not objected to by: -current
2002-01-11 15:49:02 +00:00
# install the kernel module; if the Makefile
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
# does not itself define the target install, the targets
# beforeinstall and afterinstall may also be used to cause
# actions immediately before and after the install target
# is executed.
#
kobjify. this gives us several benefits, including: * easier extensibility- new optional methods can be added to ac97/mixer/channel classes without having to fixup every driver. * forward compatibility for drivers, provided no new mandatory methods are added.
2000-12-18 01:36:41 +00:00
# load:
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# Load a module.
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
# unload:
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# Unload a module.
Add some comments for variables and targets.
1996-05-25 23:09:49 +00:00
#
In the unload target, check that the module is loaded first. Add a reload target which unloads and then loads the module.
2016-01-01 03:59:09 +00:00
# reload:
# Unload if loaded, then load.
#
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
Use vnode_if.awk rather than vnode_if.pl
2002-03-01 01:21:29 +00:00
AWK?= awk
1) s/mod(un)?load/kld$1load/ 2) s/MODLOAD/KMODLOAD/ to be consistent with the rest of the variables (KMOD, KMODOWN, KMODGRP, etc) and definition of MODLOAD/UNLOAD in the Makefile of the ATAPI module 3) textual fixups
1999-09-06 20:11:59 +00:00
KMODLOAD?= /sbin/kldload
KMODUNLOAD?= /sbin/kldunload
In the unload target, check that the module is loaded first. Add a reload target which unloads and then loads the module.
2016-01-01 03:59:09 +00:00
KMODISLOADED?= /sbin/kldstat -q -n
When building a debugging kernel with modules, build modules with debugging support as well. Debugging module support is handled identically to kernel debugging support, right down to poor choice of make variable names.
2001-08-02 09:22:18 +00:00
OBJCOPY?= objcopy
add variables MODLOAD and MODUNLOAD for modlad/modunload commands
1996-10-06 22:10:35 +00:00
Use standard bsd.init.mk prologue.
2002-04-22 15:47:11 +00:00
.include <bsd.init.mk>
Include the kernel options after bsd.init.mk so that when we're compiling on older systems that define all the MK variables in bsd.own.mk we don't get the multiply defined error for options that live in both.
2014-08-04 22:34:56 +00:00
# Grab all the options for a kernel build. For backwards compat, we need to
# do this after bsd.own.mk.
.include "kern.opts.mk"
Introduce a new make variable COMPILER_TYPE that specifies what type of compiler is being used (currently clang or gcc). COMPILER_TYPE is set in the new bsd.compiler.mk file based on the value of the CC variable or, should it prove informative, by running ${CC} --version and examining the output. To avoid negative performance impacts in the default case and correct value for COMPILER_TYPE type is determined and passed in the environment of submake instances while building world. Replace adhoc attempts at determining the compiler type by examining CC or MK_CLANG_IS_CC with checks of COMPILER_TYPE. This eliminates bootstrapping complications when first setting WITH_CLANG_IS_CC. Sponsored by: DARPA, AFRL Reviewed by: Yamaya Takashi <yamayan@kbh.biglobe.ne.jp>, imp, linimon (with some modifications post review) MFC after: 2 weeks
2012-09-13 16:00:46 +00:00
.include <bsd.compiler.mk>
Populate new KERN_OPTS from all the opt_*.h files in KERNBUILDDIR. Come up with some sensible defaults (though listing them in kmod.mk may be unwise -- we have no easy way to know what are the best sensible defaults for everything so we just catch the big stuff). Append SRCS.${opt} for each option in KERN_OPTS to SRCS to allow easy conditional compilation. Append any notion of KERN_OPTS_EXTRA to the list of kernel opts. Differential Revision: https://reviews.freebsd.org/D1530
2015-02-18 15:25:19 +00:00
.include "config.mk"
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
Define SYSDIR earlier since the defined(FIRMWS) block depends on it. It uses it in a dependency, which won't work if it is not yet defined. Sponsored by: EMC / Isilon Storage Division
2016-05-31 23:14:15 +00:00
# Search for kernel source tree in standard places.
.for _dir in ${.CURDIR}/../.. ${.CURDIR}/../../.. /sys /usr/src/sys
.if !defined(SYSDIR) && exists(${_dir}/kern/)
SYSDIR= ${_dir:tA}
.endif
.endfor
.if !defined(SYSDIR) || !exists(${SYSDIR}/kern/)
.error "can't find kernel source tree"
.endif
Kill MFILES and find things automatically. It turned out to be only lightly used. Find the proper .m file when we depend on *_if.[ch] in the srcs line, with seat-belts for false positive matches. This uses make's path mechanism. A further refinement would be to calculate this once, and then pass the resulting _MPATH to modules submakes. Differential Revision: https://reviews.freebsd.org/D2327
2015-07-03 01:50:26 +00:00
.SUFFIXES: .out .o .c .cc .cxx .C .y .l .s .S .m
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
Use MACHINE_CPUARCH as appropriate Define __KLD_SHARED to be yes or no depending on if the target uses shared binaries for klds or not (this also eliminates 4 uses of MACHINE_ARCH).
2010-09-13 07:16:48 +00:00
# amd64 and mips use direct linking for kmod, all others use shared binaries
.if ${MACHINE_CPUARCH} != amd64 && ${MACHINE_CPUARCH} != mips
__KLD_SHARED=yes
.else
__KLD_SHARED=no
.endif
Remove support for the Intel C Compiler from the build infrastructure. This support has not worked for several years, and is not likely to work again, unless Intel decides to release a native FreeBSD version of their compiler. ;)
2011-04-19 18:09:21 +00:00
.if !empty(CFLAGS:M-O[23s]) && empty(CFLAGS:M-fno-strict-aliasing)
Further embellish rev. 1.180: protect against -fno-strict-aliasing being already in CFLAGS (as is the case with "make buildkernel"). NB: rev. 1.180 is only needed due to a broken setting of CFLAGS in tinderbox.
2005-02-11 21:04:20 +00:00
CFLAGS+= -fno-strict-aliasing
This are the build infrastructure changes to allow to use the Intel C/C++ compiler (lang/icc) to build the kernel. The icc CPUTYPE CFLAGS use icc v7 syntax, icc v8 moans about them, but doesn't abort. They also produce CPU specific code (new instructions of the CPU, not only CPU specific scheduling), so if you get coredumps with signal 4 (SIGILL, illegal instruction) you've used the wrong CPUTYPE. Incarnations of this patch survive gcc compiles and my make universe. I use it on my desktop. To use it update share/mk, add /usr/local/intel/compiler70/ia32/bin (icc v7, works) or /usr/local/intel_cc_80/bin (icc v8, doesn't work) to your PATH, make sure you have a new kernel compile directory (e.g. MYKERNEL_icc) and run CFLAGS="-O2 -ip" CC=icc make depend CFLAGS="-O2 -ip" CC=icc make in it. Don't compile with -ipo, the build infrastructure uses ld directly to link the kernel and the modules, but -ipo needs the link step to be performed with Intel's linker. Problems with icc v8: - panic: npx0 cannot be emulated on an SMP system - UP: first start of /bin/sh results in a FP exception Parts of this commit contains suggestions or submissions from Marius Strobl <marius@alchemy.franken.de>. Reviewed by: silence on -arch Submitted by: netchild
2004-03-12 21:36:12 +00:00
.endif
Remove support for the Intel C Compiler from the build infrastructure. This support has not worked for several years, and is not likely to work again, unless Intel decides to release a native FreeBSD version of their compiler. ;)
2011-04-19 18:09:21 +00:00
WERROR?= -Werror
Generally force -Werror for modules when not compiling with icc. This has burned me for the last time.
2005-01-26 15:49:23 +00:00
CFLAGS+= ${WERROR}
Removed COPTS support from kmod.mk and kern.pre.mk. COPTS support in bsd.prog.mk is preserved but discouraged.
2004-08-13 14:30:26 +00:00
CFLAGS+= -D_KERNEL
Define KLD_MODULE if building a KLD module. Forgot to mention that the previous commit was Submitted by: Doug Rabson <dfr@freebsd.org>
1998-10-02 05:08:09 +00:00
CFLAGS+= -DKLD_MODULE
Generate symlinks to the "sys" and <machine> directories and put them in the include path. This fixes recent breakage of the syscons LKMs and general brokenness of the include paths (headers under /usr/include were used in many cases).
1998-01-26 20:36:38 +00:00
Restrict the include search path using -nostdinc -I- in $(CFLAGS} as in kernel Makefiles. Nothing in /usr/include is used (provided relative paths for sys/* and <machine> can be found), so there is no need for the -I/usr/include kludge as in kernel Makefiles.
1998-02-01 17:19:54 +00:00
# Don't use any standard or source-relative include directories.
This are the build infrastructure changes to allow to use the Intel C/C++ compiler (lang/icc) to build the kernel. The icc CPUTYPE CFLAGS use icc v7 syntax, icc v8 moans about them, but doesn't abort. They also produce CPU specific code (new instructions of the CPU, not only CPU specific scheduling), so if you get coredumps with signal 4 (SIGILL, illegal instruction) you've used the wrong CPUTYPE. Incarnations of this patch survive gcc compiles and my make universe. I use it on my desktop. To use it update share/mk, add /usr/local/intel/compiler70/ia32/bin (icc v7, works) or /usr/local/intel_cc_80/bin (icc v8, doesn't work) to your PATH, make sure you have a new kernel compile directory (e.g. MYKERNEL_icc) and run CFLAGS="-O2 -ip" CC=icc make depend CFLAGS="-O2 -ip" CC=icc make in it. Don't compile with -ipo, the build infrastructure uses ld directly to link the kernel and the modules, but -ipo needs the link step to be performed with Intel's linker. Problems with icc v8: - panic: npx0 cannot be emulated on an SMP system - UP: first start of /bin/sh results in a FP exception Parts of this commit contains suggestions or submissions from Marius Strobl <marius@alchemy.franken.de>. Reviewed by: silence on -arch Submitted by: netchild
2004-03-12 21:36:12 +00:00
NOSTDINC= -nostdinc
Disable -Werror for now. Remove -I- construct obsolete in GCC 4.2.
2007-05-19 04:48:08 +00:00
CFLAGS:= ${CFLAGS:N-I*} ${NOSTDINC} ${INCLMAGIC} ${CFLAGS:M-I*}
Include opt_global.h in the modules build, when building from a normal kernel build. This makes it possible for me not to get pissed off that random.ko crashes the system trying to rdtsc() when the i386/cpu.h support code decides it's okay to call that op when neither I386_CPU or I486_CPU is defined. I guess it also makes WITNESS/INVARIANTS defines get picked up by the modules.
2003-11-14 16:04:11 +00:00
.if defined(KERNBUILDDIR)
Define HAVE_KERNEL_OPTION_HEADERS when building kernel and when building modules along with kernel. After this change it is possible to embrace opt_*.h includes with ifdef HAVE_KERNEL_OPTION_HEADERS. And thus, avoid editing a lot of Makefiles in modules directory each time we introduce a new opt_xxx.h. Requested by: bde
2005-10-05 10:05:55 +00:00
CFLAGS+= -DHAVE_KERNEL_OPTION_HEADERS -include ${KERNBUILDDIR}/opt_global.h
Include opt_global.h in the modules build, when building from a normal kernel build. This makes it possible for me not to get pissed off that random.ko crashes the system trying to rdtsc() when the i386/cpu.h support code decides it's okay to call that op when neither I386_CPU or I486_CPU is defined. I guess it also makes WITNESS/INVARIANTS defines get picked up by the modules.
2003-11-14 16:04:11 +00:00
.endif
Restrict the include search path using -nostdinc -I- in $(CFLAGS} as in kernel Makefiles. Nothing in /usr/include is used (provided relative paths for sys/* and <machine> can be found), so there is no need for the -I/usr/include kludge as in kernel Makefiles.
1998-02-01 17:19:54 +00:00
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
# Add -I paths for system headers. Individual module makefiles don't
Restrict the include search path using -nostdinc -I- in $(CFLAGS} as in kernel Makefiles. Nothing in /usr/include is used (provided relative paths for sys/* and <machine> can be found), so there is no need for the -I/usr/include kludge as in kernel Makefiles.
1998-02-01 17:19:54 +00:00
# need any -I paths for this. Similar defaults for .PATH can't be
# set because there are no standard paths for non-headers.
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
CFLAGS+= -I. -I${SYSDIR}
Generate symlinks to the "sys" and <machine> directories and put them in the include path. This fixes recent breakage of the syscons LKMs and general brokenness of the include paths (headers under /usr/include were used in many cases).
1998-01-26 20:36:38 +00:00
Simplify clang ifdefs in the kernel a bit. Introduce CFLAGS.${COMPILER_TYPE} to mirror userland. Be explicit about which compiler needs something (not clang isn't necessarily gcc in the future).
2014-05-10 16:38:09 +00:00
CFLAGS.gcc+= -finline-limit=${INLINE_LIMIT}
Globally enable -fms-extensions when building kernel with gcc, and remove this option from all modules that enable it theirselves. In C mode -fms-extensions option enables anonymous structs and unions, allowing us to use this C11 feature in kernel. Of course, clang supports it without any extra options. Reviewed by: dim
2015-02-17 19:27:14 +00:00
CFLAGS.gcc+= -fms-extensions
Simplify clang ifdefs in the kernel a bit. Introduce CFLAGS.${COMPILER_TYPE} to mirror userland. Be explicit about which compiler needs something (not clang isn't necessarily gcc in the future).
2014-05-10 16:38:09 +00:00
CFLAGS.gcc+= --param inline-unit-growth=100
CFLAGS.gcc+= --param large-function-growth=1000
Provide a knob for raising the inline instruction estimate threshold and set an initial value. This is aimed at getting us closer to being able to turn -Werror back on and we can adjust the settings later on. Yes, we could turn off -Wno-inline instead, but that would hide the effect of gcc's bogo-estimator ignoring inline (either rightly or wrongly).
2003-07-23 20:03:42 +00:00
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
# Disallow common variables, and if we end up with commons from
# somewhere unexpected, allocate storage for them in the module itself.
CFLAGS+= -fno-common
LDFLAGS+= -d -warn-common
Introduce LINKER_FEATURES to avoid duplicating version logic Submitted by: bdrewery Reported by: kib
2017-06-23 17:21:37 +00:00
.if ${LINKER_FEATURES:Mbuild-id}
enable --build-id for the kernel link A Build-ID is an identifier generated at link time to uniquely identify ELF binaries. It allows efficient confirmation that an executable or shared library and a corresponding standalone debuginfo file match. (Otherwise, a checksum of the debuginfo file must be calculated when opening it in a debugger.) The FreeBSD base system includes GNU bfd ld 2.17.50 as the linker for architectures other than arm64. Build-ID support was added to bfd ld shortly after that version, so was not previously available to us. We can now start making use of Build-ID as we migrate to using lld or bfd ld from ports, conditionally enabled based on the LINKER_TYPE and LINKER_VERSION make variables added in r320244 and subsequent commits. Reviewed by: dim MFC after: 3 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D11314
2017-06-23 15:57:58 +00:00
LDFLAGS+= -Wl,--build-id=sha1
.endif
Add ${DEBUG_FLAGS} to CFLAGS in bsd.kmod.mk, not in scattered module makefiles. Bad examples in fxp/Makefile keep getting copied to new makefiles.
2000-01-28 11:26:46 +00:00
CFLAGS+= ${DEBUG_FLAGS}
Use MACHINE_CPUARCH as appropriate Define __KLD_SHARED to be yes or no depending on if the target uses shared binaries for klds or not (this also eliminates 4 uses of MACHINE_ARCH).
2010-09-13 07:16:48 +00:00
.if ${MACHINE_CPUARCH} == amd64
amd64: use -mno-omit-leaf-frame-pointer in addition to -fno-omit-frame-pointer ... in kernel builds. This is to make behavior of clang consistent with behavior of gcc. MFC after: 2 weeks
2013-07-09 08:42:02 +00:00
CFLAGS+= -fno-omit-frame-pointer -mno-omit-leaf-frame-pointer
Unconditionally add `-fno-omit-frame-pointer' to CFLAGS when building modules on the amd64 platform. Without this, ddb stack traces cannot follow module function calls, which makes debugging very difficult.
2004-08-29 02:00:50 +00:00
.endif
Add ${DEBUG_FLAGS} to CFLAGS in bsd.kmod.mk, not in scattered module makefiles. Bad examples in fxp/Makefile keep getting copied to new makefiles.
2000-01-28 11:26:46 +00:00
Build riscv modules as PIC. Submitted by: Yukishige Shibata <y-shibat@mtd.biglobe.ne.jp>
2016-06-01 14:05:32 +00:00
.if ${MACHINE_CPUARCH} == "aarch64" || ${MACHINE_CPUARCH} == "riscv"
Build changes that allow the modules on arm64. - Move the required kernel compiler flags from Makefile.arm64 to kern.mk. - Build arm64 modules as PIC; non-PIC relocations in .o for shared object output cannot be handled. - Do not try to install aarch64 symlink. - A hack for arm64 to avoid ld -r stage. See the comment for the explanation. Some functionality is lost, like ctf handling, but hopefully will be restored after newer linker is available. Reviewed by: andrew, emaste Tested by: andrew (on real hardware) Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3796
2015-10-08 17:42:08 +00:00
CFLAGS+= -fPIC
.endif
Put in a workaround for bug 196407 (arm modules cause crashes & panics). When we started compiling the kernel with -march=armv7 the compiler started emitting new types of relocation info which are incompatible with the shared-lib file format used by .ko modules. This workaround prevents the compiler from emitting the instruction sequences that require the new relocs. This amounts to using an undocumented internal compiler flag, so this is just a temporary workaround while we look for a good fix. PR: 196407
2015-01-02 03:20:13 +00:00
# Temporary workaround for PR 196407, which contains the fascinating details.
Don't allow clang to use fpu instructions or registers in kernel modules. Submitted by: Keith White <kwhite@site.uottawa.ca>
2015-01-04 00:12:30 +00:00
# Don't allow clang to use fpu instructions or registers in kernel modules.
Put in a workaround for bug 196407 (arm modules cause crashes & panics). When we started compiling the kernel with -march=armv7 the compiler started emitting new types of relocation info which are incompatible with the shared-lib file format used by .ko modules. This workaround prevents the compiler from emitting the instruction sequences that require the new relocs. This amounts to using an undocumented internal compiler flag, so this is just a temporary workaround while we look for a good fix. PR: 196407
2015-01-02 03:20:13 +00:00
.if ${MACHINE_CPUARCH} == arm
For clang 3.8.0 and higher targeting arm, use the new -mno-movt flag to disable emitting movt instructions.
2016-01-07 22:44:58 +00:00
.if ${COMPILER_VERSION} < 30800
Put in a workaround for bug 196407 (arm modules cause crashes & panics). When we started compiling the kernel with -march=armv7 the compiler started emitting new types of relocation info which are incompatible with the shared-lib file format used by .ko modules. This workaround prevents the compiler from emitting the instruction sequences that require the new relocs. This amounts to using an undocumented internal compiler flag, so this is just a temporary workaround while we look for a good fix. PR: 196407
2015-01-02 03:20:13 +00:00
CFLAGS.clang+= -mllvm -arm-use-movt=0
For clang 3.8.0 and higher targeting arm, use the new -mno-movt flag to disable emitting movt instructions.
2016-01-07 22:44:58 +00:00
.else
CFLAGS.clang+= -mno-movt
.endif
Don't allow clang to use fpu instructions or registers in kernel modules. Submitted by: Keith White <kwhite@site.uottawa.ca>
2015-01-04 00:12:30 +00:00
CFLAGS.clang+= -mfpu=none
Allow the ARM unwinder to work through modules. This will be used to add support for unwinding from dtrace. Tested by: gnn (with dtrace) Sponsored by: ABT Systems Ltd
2015-02-19 12:06:57 +00:00
CFLAGS+= -funwind-tables
Put in a workaround for bug 196407 (arm modules cause crashes & panics). When we started compiling the kernel with -march=armv7 the compiler started emitting new types of relocation info which are incompatible with the shared-lib file format used by .ko modules. This workaround prevents the compiler from emitting the instruction sequences that require the new relocs. This amounts to using an undocumented internal compiler flag, so this is just a temporary workaround while we look for a good fix. PR: 196407
2015-01-02 03:20:13 +00:00
.endif
Use MACHINE_CPUARCH as appropriate Define __KLD_SHARED to be yes or no depending on if the target uses shared binaries for klds or not (this also eliminates 4 uses of MACHINE_ARCH).
2010-09-13 07:16:48 +00:00
.if ${MACHINE_CPUARCH} == powerpc
Don't lose the frame pointer for PPC modules: backtrace doesn't work.
2005-02-02 23:44:24 +00:00
CFLAGS+= -mlongcall -fno-omit-frame-pointer
Modules on PPC need to be compiled with -mlongcall to get around the +/-64k blr offset limitation. With gcc bug #12769 fixed, it's time to put enable this.
2005-01-06 06:26:11 +00:00
.endif
Use MACHINE_CPUARCH as appropriate Define __KLD_SHARED to be yes or no depending on if the target uses shared binaries for klds or not (this also eliminates 4 uses of MACHINE_ARCH).
2010-09-13 07:16:48 +00:00
.if ${MACHINE_CPUARCH} == mips
Kernel module support for mips. Reviewed by: gonzo Tested by: Alexandr Rybalko (ray@dlink.ua)
2010-02-18 05:49:52 +00:00
CFLAGS+= -G0 -fno-pic -mno-abicalls -mlong-calls
.endif
Do the ctfmerge (= all symbols, not only symbols from one object file) for KLDs too. Noticed by: np
2010-04-22 09:24:01 +00:00
.if defined(DEBUG) || defined(DEBUG_FLAGS)
CTFFLAGS+= -g
.endif
firmware(9) is a subsystem to load binary data into the kernel via a specially crafted module. There are several handrolled sollutions to this problem in the tree already which will be replaced with this. They include iwi(4), ipw(4), ispfw(4) and digi(4). No objection from: arch MFC after: 2 weeks X-MFC after: some drivers have been converted
2006-01-29 02:52:42 +00:00
.if defined(FIRMWS)
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${KMOD:S/$/.c/}: ${SYSDIR}/tools/fw_stub.awk
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/fw_stub.awk ${FIRMWS} -m${KMOD} -c${KMOD:S/$/.c/g} \
- Add Intel firmwares for Intel PRO/Wireless LAN 2100/2200/2915 cards in a uuencoded format along with their respective LICENSE files. - Add new share/doc/legal directory to BSD.usr.dist mtree file. This is the place we install LICENSE files for restricted firmwares. - Teach firmware(9) and kmod.mk about licensed firmwares. Restricted firmwares won't load properly unless legal.<name>.license_ack is set to 1, either via kenv(1) or /boot/loader.conf. Reviewed by: mlaier, sam Permitted by: Intel (via Andrew Wilson) MFC after: 1 month
2007-03-02 11:42:56 +00:00
${FIRMWARE_LICENSE:C/.+/-l/}${FIRMWARE_LICENSE}
firmware(9) is a subsystem to load binary data into the kernel via a specially crafted module. There are several handrolled sollutions to this problem in the tree already which will be replaced with this. They include iwi(4), ipw(4), ispfw(4) and digi(4). No objection from: arch MFC after: 2 weeks X-MFC after: some drivers have been converted
2006-01-29 02:52:42 +00:00
SRCS+= ${KMOD:S/$/.c/}
CLEANFILES+= ${KMOD:S/$/.c/}
.for _firmw in ${FIRMWS}
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${_firmw:C/\:.*$/.fwo/:T}: ${_firmw:C/\:.*$//}
firmware(9) is a subsystem to load binary data into the kernel via a specially crafted module. There are several handrolled sollutions to this problem in the tree already which will be replaced with this. They include iwi(4), ipw(4), ispfw(4) and digi(4). No objection from: arch MFC after: 2 weeks X-MFC after: some drivers have been converted
2006-01-29 02:52:42 +00:00
@${ECHO} ${_firmw:C/\:.*$//} ${.ALLSRC:M*${_firmw:C/\:.*$//}}
${.CURDIR} != pwd and since we use the shell version of if now we don't even need to prepend it. Found-by: gallatin
2006-02-20 01:08:33 +00:00
@if [ -e ${_firmw:C/\:.*$//} ]; then \
Reviewed by: imp LDFLAGS is supposed to be given to CC not LD. Define _LDFLAGS as a filtered version of LDFLAGS safe to give to LD
2014-07-26 04:38:09 +00:00
${LD} -b binary --no-warn-mismatch ${_LDFLAGS} \
Always pass -m to ld for converting binary files to kernel ELF objects This is in preparation for linking with LLVM's lld, which does not have a compiled-in default output emulation. lld requires that it is specified via the -m option, or obtained from the object file(s) being linked. This will also allow all build targets to share a common linker binary. Reviewed by: imp Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D7837
2016-09-20 17:07:14 +00:00
-m ${LD_EMULATION} -r -d \
-o ${.TARGET} ${_firmw:C/\:.*$//}; \
Use shell's version of if to check if the firmware really exists in the current directory to allow user rules to create the firmware (e.g. from a uuencoded blob). make's version of if is evaluated too early to catch this. Found-by: gallatin
2006-02-18 01:38:44 +00:00
else \
ln -s ${.ALLSRC:M*${_firmw:C/\:.*$//}} ${_firmw:C/\:.*$//}; \
Reviewed by: imp LDFLAGS is supposed to be given to CC not LD. Define _LDFLAGS as a filtered version of LDFLAGS safe to give to LD
2014-07-26 04:38:09 +00:00
${LD} -b binary --no-warn-mismatch ${_LDFLAGS} \
Always pass -m to ld for converting binary files to kernel ELF objects This is in preparation for linking with LLVM's lld, which does not have a compiled-in default output emulation. lld requires that it is specified via the -m option, or obtained from the object file(s) being linked. This will also allow all build targets to share a common linker binary. Reviewed by: imp Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D7837
2016-09-20 17:07:14 +00:00
-m ${LD_EMULATION} -r -d \
-o ${.TARGET} ${_firmw:C/\:.*$//}; \
Use shell's version of if to check if the firmware really exists in the current directory to allow user rules to create the firmware (e.g. from a uuencoded blob). make's version of if is evaluated too early to catch this. Found-by: gallatin
2006-02-18 01:38:44 +00:00
rm ${_firmw:C/\:.*$//}; \
fi
firmware(9) is a subsystem to load binary data into the kernel via a specially crafted module. There are several handrolled sollutions to this problem in the tree already which will be replaced with this. They include iwi(4), ipw(4), ispfw(4) and digi(4). No objection from: arch MFC after: 2 weeks X-MFC after: some drivers have been converted
2006-01-29 02:52:42 +00:00
Don't store generated firmware object files in the source directory. Trim the leading directory of a firmware source file from the resulting target object file name so the object file is stored in the object directory. Previously, using 'FIRMWS= /path/to/fw.bin:fw.bin' would store the generated 'fw.bin.fwo' file in the /path/to directory. Now it stores it in the object directory of the kernel module being built. Reviewed by: bdrewery MFC after: 1 week Differential Revision: https://reviews.freebsd.org/D6285
2016-05-10 03:34:20 +00:00
OBJS+= ${_firmw:C/\:.*$/.fwo/:T}
firmware(9) is a subsystem to load binary data into the kernel via a specially crafted module. There are several handrolled sollutions to this problem in the tree already which will be replaced with this. They include iwi(4), ipw(4), ispfw(4) and digi(4). No objection from: arch MFC after: 2 weeks X-MFC after: some drivers have been converted
2006-01-29 02:52:42 +00:00
.endfor
.endif
Populate new KERN_OPTS from all the opt_*.h files in KERNBUILDDIR. Come up with some sensible defaults (though listing them in kmod.mk may be unwise -- we have no easy way to know what are the best sensible defaults for everything so we just catch the big stuff). Append SRCS.${opt} for each option in KERN_OPTS to SRCS to allow easy conditional compilation. Append any notion of KERN_OPTS_EXTRA to the list of kernel opts. Differential Revision: https://reviews.freebsd.org/D1530
2015-02-18 15:25:19 +00:00
# Conditionally include SRCS based on kernel config options.
.for _o in ${KERN_OPTS}
SRCS+=${SRCS.${_o}}
.endfor
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
OBJS+= ${SRCS:N*.h:R:S/$/.o/g}
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
.if !defined(PROG)
Changes to support building of KLD modules. This includes the possibly to be removed KMODDEPS define.
1998-10-02 04:51:10 +00:00
PROG= ${KMOD}.ko
.endif
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
Reduce the number of knobs controlling the build of debug modules to one, DEBUG_FLAGS, which is also compatible with <bsd.prog.mk>. Previously one had to set both DEBUG and DEBUG_FLAGS to build the .ko.debug with debugging symbols which was boring when doing this manually.
2004-02-13 10:40:54 +00:00
.if !defined(DEBUG_FLAGS)
When building a debugging kernel with modules, build modules with debugging support as well. Debugging module support is handled identically to kernel debugging support, right down to poor choice of make variable names.
2001-08-02 09:22:18 +00:00
FULLPROG= ${PROG}
.else
Install kernel debug data under /usr/lib/debug This avoids needing a large boot partition / file system in order to accommodate multiple kernels, and provides consistency with userland debug. This also simplifies the process of moving kernel debug files to a separate package and installing them on demand. In addition, change kernel debug file extension to .debug, to match userland debug files. When using the supported kernel installation method the /usr/lib/debug/boot/kernel directory will be renamed (to kernel.old) as is done with /boot/kernel. Developers wishing to maintain the historical behavior of installing debug files in /boot/kernel/ can set KERN_DEBUGDIR="" in src.conf(5). Reviewed by: bdrewery, brooks, imp, markj Relnotes: yes Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D1006
2015-09-24 16:55:22 +00:00
FULLPROG= ${PROG}.full
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${PROG}: ${FULLPROG} ${PROG}.debug
Install kernel debug data under /usr/lib/debug This avoids needing a large boot partition / file system in order to accommodate multiple kernels, and provides consistency with userland debug. This also simplifies the process of moving kernel debug files to a separate package and installing them on demand. In addition, change kernel debug file extension to .debug, to match userland debug files. When using the supported kernel installation method the /usr/lib/debug/boot/kernel directory will be renamed (to kernel.old) as is done with /boot/kernel. Developers wishing to maintain the historical behavior of installing debug files in /boot/kernel/ can set KERN_DEBUGDIR="" in src.conf(5). Reviewed by: bdrewery, brooks, imp, markj Relnotes: yes Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D1006
2015-09-24 16:55:22 +00:00
${OBJCOPY} --strip-debug --add-gnu-debuglink=${PROG}.debug \
Installing debug modules was a bad idea -- I bogusly assumed that our kernel linker will only load PT_LOAD segments, apparently not. Instead, produce .dbg objects from .debug objects, and install them together with non-debug objects, as described in objcopy(1). Original code by: obrien
2005-10-27 14:24:45 +00:00
${FULLPROG} ${.TARGET}
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${PROG}.debug: ${FULLPROG}
Installing debug modules was a bad idea -- I bogusly assumed that our kernel linker will only load PT_LOAD segments, apparently not. Instead, produce .dbg objects from .debug objects, and install them together with non-debug objects, as described in objcopy(1). Original code by: obrien
2005-10-27 14:24:45 +00:00
${OBJCOPY} --only-keep-debug ${FULLPROG} ${.TARGET}
When building a debugging kernel with modules, build modules with debugging support as well. Debugging module support is handled identically to kernel debugging support, right down to poor choice of make variable names.
2001-08-02 09:22:18 +00:00
.endif
Use MACHINE_CPUARCH as appropriate Define __KLD_SHARED to be yes or no depending on if the target uses shared binaries for klds or not (this also eliminates 4 uses of MACHINE_ARCH).
2010-09-13 07:16:48 +00:00
.if ${__KLD_SHARED} == yes
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${FULLPROG}: ${KMOD}.kld
add -znotext to kernel module link invocation ARM kernel modules require .text relocations (DT_TEXTREL) in shared object ouptut, which is not allowed by default by lld. Add the -znotext option to enable this. For simplicity add it unconditionally: it is already default and thus either redundant (GNU BFD ld and gold from ports) or ignored as an unknown option (GNU BFD ld 2.17.50 in the base system). Reviewed by: kib MFC after: 3 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D11250
2017-06-21 00:33:16 +00:00
${LD} -m ${LD_EMULATION} -Bshareable -znotext ${_LDFLAGS} \
-o ${.TARGET} ${KMOD}.kld
Reduce the number of knobs controlling the build of debug modules to one, DEBUG_FLAGS, which is also compatible with <bsd.prog.mk>. Previously one had to set both DEBUG and DEBUG_FLAGS to build the .ko.debug with debugging symbols which was boring when doing this manually.
2004-02-13 10:40:54 +00:00
.if !defined(DEBUG_FLAGS)
${OBJCOPY} --strip-debug ${.TARGET}
We didn't strip debugging symbols from .ko if DEBUG was undefined.
2004-02-13 09:39:45 +00:00
.endif
Enable first part of kld's on amd64. This is known to not work right yet, but building kld's is OK now and they can be loaded by kldload(2). (but the machine will likely crash soon afterwards, a "minor" problem :-) Brought to you by: my injured knee (from moving)
2004-05-16 20:11:38 +00:00
.endif
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
Reverse the sense of EXPORT_SYMS. If EXPORT_SYMS is not defined, no symbols are exported from the module. This is the typical configuration for most device drivers and standalone modules; only infrastructure modules or those with special requirements typically need to export symbols. Don't print the objcopy commands as they are run when converting symbols; they're bulky and annoying in many cases.
2002-01-11 01:11:44 +00:00
EXPORT_SYMS?= NO
.if ${EXPORT_SYMS} != YES
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
CLEANFILES+= export_syms
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
.endif
Use MACHINE_CPUARCH as appropriate Define __KLD_SHARED to be yes or no depending on if the target uses shared binaries for klds or not (this also eliminates 4 uses of MACHINE_ARCH).
2010-09-13 07:16:48 +00:00
.if ${__KLD_SHARED} == yes
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${KMOD}.kld: ${OBJS}
Enable first part of kld's on amd64. This is known to not work right yet, but building kld's is OK now and they can be loaded by kldload(2). (but the machine will likely crash soon afterwards, a "minor" problem :-) Brought to you by: my injured knee (from moving)
2004-05-16 20:11:38 +00:00
.else
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${FULLPROG}: ${OBJS}
Enable first part of kld's on amd64. This is known to not work right yet, but building kld's is OK now and they can be loaded by kldload(2). (but the machine will likely crash soon afterwards, a "minor" problem :-) Brought to you by: my injured knee (from moving)
2004-05-16 20:11:38 +00:00
.endif
Always pass the linker emulation via -m when linking modules and kernels. Previously the linker emulation was only passed when building binary objects for firmware modules. This change always passes the desired output format for kernel modules and kernels rather than requiring the toolchain's default output format to match the desired output format. This in turn permits use of external toolchains whose default output format does not match the desired output format. Reviewed by: imp, emaste Sponsored by: DARPA / AFRL Differential Revision: https://reviews.freebsd.org/D10085
2017-04-05 03:05:04 +00:00
${LD} -m ${LD_EMULATION} ${_LDFLAGS} -r -d -o ${.TARGET} ${OBJS}
bitrotted compat cruft removal: o KMODDEPS warning is 15 years stale. Remove it. o MK_CTF will always be defined now, so no need to test to see if it is defined. o no need to define MK_FORMAT_EXTENTIONS if undefined anymore.
2014-05-10 16:39:15 +00:00
.if ${MK_CTF} != "no"
Conditionalize ctfconvert/ctfmerge runs on make level (.if/.endif) instead of executing a shell on every object or executable/library file. This shaves off more than 30,000 shell invocations during buildworld.
2011-11-29 08:38:47 +00:00
${CTFMERGE} ${CTFFLAGS} -o ${.TARGET} ${OBJS}
.endif
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
.if defined(EXPORT_SYMS)
Reverse the sense of EXPORT_SYMS. If EXPORT_SYMS is not defined, no symbols are exported from the module. This is the typical configuration for most device drivers and standalone modules; only infrastructure modules or those with special requirements typically need to export symbols. Don't print the objcopy commands as they are run when converting symbols; they're bulky and annoying in many cases.
2002-01-11 01:11:44 +00:00
.if ${EXPORT_SYMS} != YES
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
.if ${EXPORT_SYMS} == NO
Don't use touch when what is really meant is :> (create an empty file, or truncate it if it exists) or :>> (ensure the file exists, but don't change it if it already does) Reviewed by: ru MFC after: 2 weeks
2006-08-14 13:28:53 +00:00
:> export_syms
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
.elif !exists(${.CURDIR}/${EXPORT_SYMS})
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
echo ${EXPORT_SYMS} > export_syms
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
.else
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
grep -v '^#' < ${EXPORT_SYMS} > export_syms
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
.endif
Consistently use ${AWK} instead of hard-coding the program name. MFC after: 3 days
2015-12-17 00:02:53 +00:00
${AWK} -f ${SYSDIR}/conf/kmod_syms.awk ${.TARGET} \
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
export_syms | xargs -J% ${OBJCOPY} % ${.TARGET}
Reverse the sense of EXPORT_SYMS. If EXPORT_SYMS is not defined, no symbols are exported from the module. This is the typical configuration for most device drivers and standalone modules; only infrastructure modules or those with special requirements typically need to export symbols. Don't print the objcopy commands as they are run when converting symbols; they're bulky and annoying in many cases.
2002-01-11 01:11:44 +00:00
.endif
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
.endif
Add the infrastructure to support loading multiple versions of TCP stack modules. It adds support for mangling symbols exported by a module by prepending a string to them. (This avoids overlapping symbols in the kernel linker.) It allows the use of a macro as the module name in the DECLARE_MACRO() and MACRO_VERSION() macros. It allows the code to register stack aliases (e.g. both a generic name ["default"] and version-specific name ["default_10_3p1"]). With these changes, it is trivial to compile TCP stack modules with the name defined in the Makefile and to load multiple versions of the same stack simultaneously. This functionality can be used to enable side-by-side testing of an old and new version of the same TCP stack. It also could support upgrading the TCP stack without a reboot. Reviewed by: gnn, sjg (makefiles only) Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D11086
2017-06-08 20:41:28 +00:00
.if defined(PREFIX_SYMS)
${AWK} -v prefix=${PREFIX_SYMS} -f ${SYSDIR}/conf/kmod_syms_prefix.awk \
${.TARGET} /dev/null | xargs -J% ${OBJCOPY} % ${.TARGET}
.endif
Use MACHINE_CPUARCH as appropriate Define __KLD_SHARED to be yes or no depending on if the target uses shared binaries for klds or not (this also eliminates 4 uses of MACHINE_ARCH).
2010-09-13 07:16:48 +00:00
.if !defined(DEBUG_FLAGS) && ${__KLD_SHARED} == no
Enable first part of kld's on amd64. This is known to not work right yet, but building kld's is OK now and they can be loaded by kldload(2). (but the machine will likely crash soon afterwards, a "minor" problem :-) Brought to you by: my injured knee (from moving)
2004-05-16 20:11:38 +00:00
${OBJCOPY} --strip-debug ${.TARGET}
.endif
Eliminate the use of commons in the kernel and modules, simplifying the module linking process and eliminating the risks associated with doubly-defined variables. Cases where commons were legitimately used (detection of compiled-in subsystems) have been converted to use sysinits, and any new code should use this or an equivalent practice as a matter of course. Modules can override this behaviour by substituting -fno-common out of ${CFLAGS} in cases where commons are necessary (eg. third-party object modules). Commons will be resolved and allocated space when the kld is linked as part of the module build process, so they will not pose a risk to the kernel or other modules. Provide a mechanism for controlling the export of symbols from the module namespace. The EXPORT_SYMS variable may be set in the Makefile to NO (export no symbols), a list of symbols to export, or the name of a file containing a newline-seperated list of symbols to be exported. Non-exported symbols are converted to local symbols. If EXPORT_SYMS is not set, all global symbols are currently exported. This behaviour is expected to change (to exporting no symbols) once modules have been converted. Reviewed by: peter (in principle) Obtained from: green (kmod_syms.awk)
2002-01-10 03:52:01 +00:00
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
_ILINKS=machine
Build changes that allow the modules on arm64. - Move the required kernel compiler flags from Makefile.arm64 to kern.mk. - Build arm64 modules as PIC; non-PIC relocations in .o for shared object output cannot be handled. - Do not try to install aarch64 symlink. - A hack for arm64 to avoid ld -r stage. See the comment for the explanation. Some functionality is lost, like ctf handling, but hopefully will be restored after newer linker is available. Reviewed by: andrew, emaste Tested by: andrew (on real hardware) Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3796
2015-10-08 17:42:08 +00:00
.if ${MACHINE} != ${MACHINE_CPUARCH} && ${MACHINE} != "arm64"
Convert several instances of MACHINE_ARCH to MACHINE_CPUARCH and use the correct compiler flags on 64-bit PowerPC.
2010-07-13 13:11:18 +00:00
_ILINKS+=${MACHINE_CPUARCH}
Don't forget to create new-style links for module builds. Forgotten by: imp@ Reminded by: nyan@
2005-04-03 04:56:05 +00:00
.endif
Add an x86/include directory to the kernel to hold headers that are common to amd64, i386, and pc98. The headers are installed to /usr/include/x86 during an installworld, and an 'x86' symlink is created for kernel builds similar to 'machine' so that the headers can be included as <x86/foo.h>. Reviewed by: imp
2010-11-01 17:34:04 +00:00
.if ${MACHINE_CPUARCH} == "i386" || ${MACHINE_CPUARCH} == "amd64"
_ILINKS+=x86
.endif
Clean symlinks created in sys/modules dirs when a make clean is invoked. Phabric: https://phabric.freebsd.org/D474 Reviewed by: grehan sjg
2014-07-24 00:25:25 +00:00
CLEANFILES+=${_ILINKS}
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
Consolidate common beforebuild logic. Sponsored by: EMC / Isilon Storage Division
2016-02-16 02:09:48 +00:00
all: ${PROG}
Generate symlinks to the "sys" and <machine> directories and put them in the include path. This fixes recent breakage of the syscons LKMs and general brokenness of the include paths (headers under /usr/include were used in many cases).
1998-01-26 20:36:38 +00:00
I accidently committed two 'cleandepend' when I was deciding which form was "cleaner".
2005-04-19 18:29:11 +00:00
beforedepend: ${_ILINKS}
kmod.mk: Ensure ILINKS are created before building. This allows skipping 'make depend' or running 'make clean all' without getting a flip-flopping dependency due to the exists() just below. Otherwise an error is encountered, such as: fatal error: 'machine/endian.h' file not found. Sponsored by: EMC / Isilon Storage Division
2016-02-08 20:58:52 +00:00
beforebuild: ${_ILINKS}
Move beforedepend: outside of the loop. # Note: we should remove the rm kludge soon. It fails in many cases. Submitted by: bde
2001-03-05 06:19:29 +00:00
If the symbolic links @ or machine exist, do not depend on them. This fixes the problem where if src/sys or src/sys/$MACHINE_ARCH/include changed at all, all the modules would be rebuilt. Reviewed by: bde
2001-02-23 04:49:31 +00:00
# Ensure that the links exist without depending on it when it exists which
# causes all the modules to be rebuilt when the directory pointed to changes.
.for _link in ${_ILINKS}
.if !exists(${.OBJDIR}/${_link})
kmod: Fix depending on ILINKS for tracked DEPENDOBJS. The objects that may be in the dependency graph may not match ${OBJS}. Ensure the ilink link is added as a dependency for all of them when a .depend file is missing for that objfile. MFC after: 2 weeks Sponsored by: Dell EMC Isilon
2017-03-17 18:08:00 +00:00
OBJS_DEPEND_GUESS+= ${_link}
If the symbolic links @ or machine exist, do not depend on them. This fixes the problem where if src/sys or src/sys/$MACHINE_ARCH/include changed at all, all the modules would be rebuilt. Reviewed by: bde
2001-02-23 04:49:31 +00:00
.endif
.endfor
Generate symlinks to the "sys" and <machine> directories and put them in the include path. This fixes recent breakage of the syscons LKMs and general brokenness of the include paths (headers under /usr/include were used in many cases).
1998-01-26 20:36:38 +00:00
Sync FreeBSD's bmake branch with Juniper's internal bmake branch. Requested by: Simon Gerraty <sjg@juniper.net>
2012-08-22 19:25:57 +00:00
.NOPATH: ${_ILINKS}
Use a full path to the target for make rules which create symlinks @, machine and ${MACHINE_CPUARCH}. Otherwise the presence of a file named "x86" or "x86.c" in the make path can cause problems. Submitted by: lwhsu (original version) MFC after: 1 month
2014-03-04 14:00:49 +00:00
Update the kld glue Pre-Approved by: jkh
1998-10-16 03:32:03 +00:00
${_ILINKS}:
Add support for compiling kernel modules outside of the tree. If you do not have the kernel you wish to compile against in either /usr/src/sys or /sys, then you will need to set SYSDIR to point to the sys directory of the source tree that contians the source. Also, minor tweaks to the load/unload targets from Bruce. I've had this through several make worlds, as well as using it on a daily basis for the past couple of weeks to build modules needed for testing at Timing Solutions. Reviewed and revised by: bde Work sponsored by: Timing Solutions
2000-04-08 17:20:00 +00:00
@case ${.TARGET} in \
Don't forget to create new-style links for module builds. Forgotten by: imp@ Reminded by: nyan@
2005-04-03 04:56:05 +00:00
machine) \
path=${SYSDIR}/${MACHINE}/include ;; \
Add an x86/include directory to the kernel to hold headers that are common to amd64, i386, and pc98. The headers are installed to /usr/include/x86 during an installworld, and an 'x86' symlink is created for kernel builds similar to 'machine' so that the headers can be included as <x86/foo.h>. Reviewed by: imp
2010-11-01 17:34:04 +00:00
*) \
Use a full path to the target for make rules which create symlinks @, machine and ${MACHINE_CPUARCH}. Otherwise the presence of a file named "x86" or "x86.c" in the make path can cause problems. Submitted by: lwhsu (original version) MFC after: 1 month
2014-03-04 14:00:49 +00:00
path=${SYSDIR}/${.TARGET:T}/include ;; \
Add support for compiling kernel modules outside of the tree. If you do not have the kernel you wish to compile against in either /usr/src/sys or /sys, then you will need to set SYSDIR to point to the sys directory of the source tree that contians the source. Also, minor tweaks to the load/unload targets from Bruce. I've had this through several make worlds, as well as using it on a daily basis for the past couple of weeks to build modules needed for testing at Timing Solutions. Reviewed and revised by: bde Work sponsored by: Timing Solutions
2000-04-08 17:20:00 +00:00
esac ; \
Update the kld glue Pre-Approved by: jkh
1998-10-16 03:32:03 +00:00
path=`(cd $$path && /bin/pwd)` ; \
Use a full path to the target for make rules which create symlinks @, machine and ${MACHINE_CPUARCH}. Otherwise the presence of a file named "x86" or "x86.c" in the make path can cause problems. Submitted by: lwhsu (original version) MFC after: 1 month
2014-03-04 14:00:49 +00:00
${ECHO} ${.TARGET:T} "->" $$path ; \
WITH_META_MDE: Fix machine/include and x86/include issues. - Fixes 'ln: File exists' errors. - Fixes creating include directories in the source directory as well. Sponsored by: EMC / Isilon Storage Division
2016-05-31 23:40:21 +00:00
ln -fhs $$path ${.TARGET:T}
Generate symlinks to the "sys" and <machine> directories and put them in the include path. This fixes recent breakage of the syscons LKMs and general brokenness of the include paths (headers under /usr/include were used in many cases).
1998-01-26 20:36:38 +00:00
- Remove include links only when .depend is also removed, so that "make depend; make clean; make -n" works. - Preseve kernel's .depend if it already exists and its creation is interrupted. Reported/reviewed by: bde
2006-10-13 22:28:14 +00:00
CLEANFILES+= ${PROG} ${KMOD}.kld ${OBJS}
In the !DEBUG case, we were passing foo.ko two times on the rm -f command line when doing a make clean. Fix this. Reviewed by: ru
2002-11-06 17:40:13 +00:00
Reduce the number of knobs controlling the build of debug modules to one, DEBUG_FLAGS, which is also compatible with <bsd.prog.mk>. Previously one had to set both DEBUG and DEBUG_FLAGS to build the .ko.debug with debugging symbols which was boring when doing this manually.
2004-02-13 10:40:54 +00:00
.if defined(DEBUG_FLAGS)
Install kernel debug data under /usr/lib/debug This avoids needing a large boot partition / file system in order to accommodate multiple kernels, and provides consistency with userland debug. This also simplifies the process of moving kernel debug files to a separate package and installing them on demand. In addition, change kernel debug file extension to .debug, to match userland debug files. When using the supported kernel installation method the /usr/lib/debug/boot/kernel directory will be renamed (to kernel.old) as is done with /boot/kernel. Developers wishing to maintain the historical behavior of installing debug files in /boot/kernel/ can set KERN_DEBUGDIR="" in src.conf(5). Reviewed by: bdrewery, brooks, imp, markj Relnotes: yes Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D1006
2015-09-24 16:55:22 +00:00
CLEANFILES+= ${FULLPROG} ${PROG}.debug
In the !DEBUG case, we were passing foo.ko two times on the rm -f command line when doing a make clean. Fix this. Reviewed by: ru
2002-11-06 17:40:13 +00:00
.endif
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
.if !target(install)
Don't install KO's with the "schg" flag. We are way too inconsistent with our setting of the "schg" flag, and in our default install, it doesn't really offer any additional security. Reviewed by: arch@
2001-04-25 20:56:15 +00:00
_INSTALLFLAGS:= ${INSTALLFLAGS}
Add variable INSTALLFLAGS_EDIT used to remove arbitrary INSTALLFLAGS. Specifically intended for removing -fschg ("INSTALLFLAGS_EDIT=:S/schg/uchg/") this makes the NOFSCHG flag redundant. NOFSCHG will still be honoured by bsd.lib.mk but is valid for buildworld only. NOFSCHG is still implemented in the old way (ie. _not_ ".if NOFSCHG then { INSTALLFLAGS_EDIT+=:S/schg/,/ }" to emphasize the fact that NOFSCHG is only supported in a limited fashion and for buildworld. The interface and implementation are such that future use of flags such as sappnd can also be easily removed or altered (perhaps to uappnd). This commit brought to you by the letters B, D, and E, and the numbers six, one, thirteen, and three.
1999-07-31 20:27:33 +00:00
.for ie in ${INSTALLFLAGS_EDIT}
_INSTALLFLAGS:= ${_INSTALLFLAGS${ie}}
.endfor
Handle installation of links through bsd.links.mk. Removed comments that no longer directly apply here.
2002-07-17 08:21:50 +00:00
.if !target(realinstall)
Install kernel debug data under /usr/lib/debug This avoids needing a large boot partition / file system in order to accommodate multiple kernels, and provides consistency with userland debug. This also simplifies the process of moving kernel debug files to a separate package and installing them on demand. In addition, change kernel debug file extension to .debug, to match userland debug files. When using the supported kernel installation method the /usr/lib/debug/boot/kernel directory will be renamed (to kernel.old) as is done with /boot/kernel. Developers wishing to maintain the historical behavior of installing debug files in /boot/kernel/ can set KERN_DEBUGDIR="" in src.conf(5). Reviewed by: bdrewery, brooks, imp, markj Relnotes: yes Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D1006
2015-09-24 16:55:22 +00:00
KERN_DEBUGDIR?= ${DEBUGDIR}
Handle installation of links through bsd.links.mk. Removed comments that no longer directly apply here.
2002-07-17 08:21:50 +00:00
realinstall: _kmodinstall
.ORDER: beforeinstall _kmodinstall
Add some missing .PHONY. These are relevant for WITH_META_MODE to ensure they are always reran and don't generate a .meta file. Sponsored by: EMC / Isilon Storage Division
2016-05-26 23:20:14 +00:00
_kmodinstall: .PHONY
Also tag the modules
2015-02-08 16:30:44 +00:00
${INSTALL} -T release -o ${KMODOWN} -g ${KMODGRP} -m ${KMODMODE} \
Do as r289391 did for share/mk and make installing to a non-existent directory an error. Most of these do a 'mkdir -p' or 'install -d' before installing, but add the trailing / here for consistency with the userland install. MFC after: 2 weeks X-MFC-With: r289391 Sponsored by: EMC / Isilon Storage Division
2015-10-17 05:49:07 +00:00
${_INSTALLFLAGS} ${PROG} ${DESTDIR}${KMODDIR}/
Revert r222186 per instructions for FreeBSD 10. (a 10-CURRENT share/mk is already required to build a 10-CURRENT kernel on 9-STABLE)
2012-07-03 05:01:00 +00:00
.if defined(DEBUG_FLAGS) && !defined(INSTALL_NODEBUG) && ${MK_KERNEL_SYMBOLS} != "no"
Also tag the modules
2015-02-08 16:30:44 +00:00
${INSTALL} -T debug -o ${KMODOWN} -g ${KMODGRP} -m ${KMODMODE} \
Do as r289391 did for share/mk and make installing to a non-existent directory an error. Most of these do a 'mkdir -p' or 'install -d' before installing, but add the trailing / here for consistency with the userland install. MFC after: 2 weeks X-MFC-With: r289391 Sponsored by: EMC / Isilon Storage Division
2015-10-17 05:49:07 +00:00
${_INSTALLFLAGS} ${PROG}.debug ${DESTDIR}${KERN_DEBUGDIR}${KMODDIR}/
Refactor (some more) installation of kernel and module objects. Try to make everyone happy: David (to have debug kernels installed by default), Warner (to be able to override that), and myself (for actually making it all work and to be consistent). Now, if kernel was configured for debugging (through DEBUG=-g in the kernel config file or "config -g"), doing "make install" will install debug versions of kernel and module objects with their canonical names, kernel.debug -> /boot/kernel/kernel if_fxp.ko.debug -> /boot/kernel/if_fxp.ko Installing a kernel not configured for debugging, or debug kernel with INSTALL_NODEBUG variable defined, will install non-debug kernel and module objects. Also, restore the install.debug and reinstall.debug targets that are part of the existing API (they cause some additional gdb(1) scripts to be installed).
2005-10-25 09:05:07 +00:00
.endif
Handle installation of links through bsd.links.mk. Removed comments that no longer directly apply here.
2002-07-17 08:21:50 +00:00
.include <bsd.links.mk>
Run kldxref on sparc64, it works now.
2003-01-21 05:52:48 +00:00
.if !defined(NO_XREF)
Handle installation of links through bsd.links.mk. Removed comments that no longer directly apply here.
2002-07-17 08:21:50 +00:00
afterinstall: _kldxref
.ORDER: realinstall _kldxref
.ORDER: _installlinks _kldxref
Add some missing .PHONY. These are relevant for WITH_META_MODE to ensure they are always reran and don't generate a .meta file. Sponsored by: EMC / Isilon Storage Division
2016-05-26 23:20:14 +00:00
_kldxref: .PHONY
Check that kldxref(8) exists before running it.
2002-05-14 07:49:12 +00:00
@if type kldxref >/dev/null 2>&1; then \
${ECHO} kldxref ${DESTDIR}${KMODDIR}; \
kldxref ${DESTDIR}${KMODDIR}; \
fi
Implement the long-awaited module->file cache database. A userland tool (kldxref(8)) keeps a cache of what modules and versions are inside what .ko files. I have tested this on both Alpha and i386. Submitted by: bp
2001-09-11 01:09:24 +00:00
.endif
Use comments after .endif to please make(1) with latest changes so that it's possible to build a kernel without getting flooded with thousands of warnings.
2005-04-11 12:54:03 +00:00
.endif # !target(realinstall)
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
Use comments after .endif to please make(1) with latest changes so that it's possible to build a kernel without getting flooded with thousands of warnings.
2005-04-11 12:54:03 +00:00
.endif # !target(install)
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
.if !target(load)
Add some missing .PHONY. These are relevant for WITH_META_MODE to ensure they are always reran and don't generate a .meta file. Sponsored by: EMC / Isilon Storage Division
2016-05-26 23:20:14 +00:00
load: ${PROG} .PHONY
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
${KMODLOAD} -v ${.OBJDIR}/${PROG}
A make system file for building LKMs, derived from bsd.prog.mk.
1994-09-14 21:59:28 +00:00
.endif
.if !target(unload)
Add some missing .PHONY. These are relevant for WITH_META_MODE to ensure they are always reran and don't generate a .meta file. Sponsored by: EMC / Isilon Storage Division
2016-05-26 23:20:14 +00:00
unload: .PHONY
In the unload target, check that the module is loaded first. Add a reload target which unloads and then loads the module.
2016-01-01 03:59:09 +00:00
if ${KMODISLOADED} ${PROG} ; then ${KMODUNLOAD} -v ${PROG} ; fi
.endif
.if !target(reload)
Add some missing .PHONY. These are relevant for WITH_META_MODE to ensure they are always reran and don't generate a .meta file. Sponsored by: EMC / Isilon Storage Division
2016-05-26 23:20:14 +00:00
reload: unload load .PHONY
Update the kld glue Pre-Approved by: jkh
1998-10-16 03:32:03 +00:00
.endif
Make it easier to create filesystem LKMs by doing most of the work in the Makefile. Just define VFS_LKM and you're off!
1994-09-21 03:49:59 +00:00
Fix the building of null module. In the DIAGNOSTICS case, we include opt_ddb.h. These changes expand green's work of including opt_global.h to prefer opt files in the kernel directory. Further refinement might be needed, but I think this is good. Note: While this is a step on the path to moving the meta information about modules into the config files, it doesn't actually do that. It just pulls in the opt files in a way that allows one to build 'generic' modules outside the tree.
2003-11-16 01:41:38 +00:00
.if defined(KERNBUILDDIR)
.PATH: ${KERNBUILDDIR}
Style fixes, including: - Sort kmod.mk knobs in the documentation section. - Fixed misuses of the word "KLD" which stands for "kernel ld", or "kernel linker", where kernel module is meant. - Removed redundant uses of ${.OBJDIR}. - Whitespace and indentation fixes. - CLEANFILES cleanup. - Target redefinition protection (install.debug). Submitted by: bde, ru Reviewed by: ru, bde
2004-10-12 15:04:10 +00:00
CFLAGS+= -I${KERNBUILDDIR}
Fix the building of null module. In the DIAGNOSTICS case, we include opt_ddb.h. These changes expand green's work of including opt_global.h to prefer opt files in the kernel directory. Further refinement might be needed, but I think this is good. Note: While this is a step on the path to moving the meta information about modules into the config files, it doesn't actually do that. It just pulls in the opt files in a way that allows one to build 'generic' modules outside the tree.
2003-11-16 01:41:38 +00:00
.for _src in ${SRCS:Mopt_*.h}
CLEANFILES+= ${_src}
.if !target(${_src})
${_src}:
When linking make sure it succeeds.
2006-02-17 04:59:23 +00:00
ln -sf ${KERNBUILDDIR}/${_src} ${.TARGET}
Fix the building of null module. In the DIAGNOSTICS case, we include opt_ddb.h. These changes expand green's work of including opt_global.h to prefer opt files in the kernel directory. Further refinement might be needed, but I think this is good. Note: While this is a step on the path to moving the meta information about modules into the config files, it doesn't actually do that. It just pulls in the opt files in a way that allows one to build 'generic' modules outside the tree.
2003-11-16 01:41:38 +00:00
.endif
.endfor
.else
Build and clean device interface files (foo_if.[ch]) and empty options files (opt_*.h) automatically (if they are in ${SRCS}). Clean vnode_if.[ch] automatically (if one of them is in ${SRCS}, not just if VFS_KLD is defined). There are some complications to avoid using the "@" symlink before it is built.
1999-11-28 17:52:40 +00:00
.for _src in ${SRCS:Mopt_*.h}
CLEANFILES+= ${_src}
.if !target(${_src})
${_src}:
Don't use touch when what is really meant is :> (create an empty file, or truncate it if it exists) or :>> (ensure the file exists, but don't change it if it already does) Reviewed by: ru MFC after: 2 weeks
2006-08-14 13:28:53 +00:00
:> ${.TARGET}
Build and clean device interface files (foo_if.[ch]) and empty options files (opt_*.h) automatically (if they are in ${SRCS}). Clean vnode_if.[ch] automatically (if one of them is in ${SRCS}, not just if VFS_KLD is defined). There are some complications to avoid using the "@" symlink before it is built.
1999-11-28 17:52:40 +00:00
.endif
.endfor
Fix the building of null module. In the DIAGNOSTICS case, we include opt_ddb.h. These changes expand green's work of including opt_global.h to prefer opt files in the kernel directory. Further refinement might be needed, but I think this is good. Note: While this is a step on the path to moving the meta information about modules into the config files, it doesn't actually do that. It just pulls in the opt files in a way that allows one to build 'generic' modules outside the tree.
2003-11-16 01:41:38 +00:00
.endif
Build and clean device interface files (foo_if.[ch]) and empty options files (opt_*.h) automatically (if they are in ${SRCS}). Clean vnode_if.[ch] automatically (if one of them is in ${SRCS}, not just if VFS_KLD is defined). There are some complications to avoid using the "@" symlink before it is built.
1999-11-28 17:52:40 +00:00
Fix a typo in a comment.
2010-03-26 08:47:07 +00:00
# Respect configuration-specific C flags.
Propagate CONF_CFLAGS (from makeoptions) to the module build too. Discussed with: jhb (on arch@)
2010-03-25 13:47:21 +00:00
CFLAGS+= ${CONF_CFLAGS}
Use the more readable empty() syntax.
2005-09-07 17:38:46 +00:00
.if !empty(SRCS:Mvnode_if.c)
Unbreak KLDs that use VFS by accounting for changes in the generation process for the vnode header files.
2004-12-08 03:54:32 +00:00
CLEANFILES+= vnode_if.c
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
vnode_if.c: ${SYSDIR}/tools/vnode_if.awk ${SYSDIR}/kern/vnode_if.src
${AWK} -f ${SYSDIR}/tools/vnode_if.awk ${SYSDIR}/kern/vnode_if.src -c
Unbreak KLDs that use VFS by accounting for changes in the generation process for the vnode header files.
2004-12-08 03:54:32 +00:00
.endif
Use the more readable empty() syntax.
2005-09-07 17:38:46 +00:00
.if !empty(SRCS:Mvnode_if.h)
Properly handle vnode_if.h dependencies, and don't forget to clean all files.
2004-12-08 07:23:12 +00:00
CLEANFILES+= vnode_if.h vnode_if_newproto.h vnode_if_typedef.h
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
vnode_if.h vnode_if_newproto.h vnode_if_typedef.h: ${SYSDIR}/tools/vnode_if.awk \
${SYSDIR}/kern/vnode_if.src
Properly handle vnode_if.h dependencies, and don't forget to clean all files.
2004-12-08 07:23:12 +00:00
vnode_if.h: vnode_if_newproto.h vnode_if_typedef.h
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/vnode_if.awk ${SYSDIR}/kern/vnode_if.src -h
Properly handle vnode_if.h dependencies, and don't forget to clean all files.
2004-12-08 07:23:12 +00:00
vnode_if_newproto.h:
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/vnode_if.awk ${SYSDIR}/kern/vnode_if.src -p
Properly handle vnode_if.h dependencies, and don't forget to clean all files.
2004-12-08 07:23:12 +00:00
vnode_if_typedef.h:
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/vnode_if.awk ${SYSDIR}/kern/vnode_if.src -q
Build and clean device interface files (foo_if.[ch]) and empty options files (opt_*.h) automatically (if they are in ${SRCS}). Clean vnode_if.[ch] automatically (if one of them is in ${SRCS}, not just if VFS_KLD is defined). There are some complications to avoid using the "@" symlink before it is built.
1999-11-28 17:52:40 +00:00
.endif
Make it easier to create filesystem LKMs by doing most of the work in the Makefile. Just define VFS_LKM and you're off!
1994-09-21 03:49:59 +00:00
Kill MFILES and find things automatically. It turned out to be only lightly used. Find the proper .m file when we depend on *_if.[ch] in the srcs line, with seat-belts for false positive matches. This uses make's path mechanism. A further refinement would be to calculate this once, and then pass the resulting _MPATH to modules submakes. Differential Revision: https://reviews.freebsd.org/D2327
2015-07-03 01:50:26 +00:00
# Build _if.[ch] from _if.m, and clean them when we're done.
Reduce duplicated logic from r291744. Sponsored by: EMC / Isilon Storage Division
2016-03-12 22:21:14 +00:00
# __MPATH defined in config.mk
Restore cleaning of auto-generated kobj files after the recent MFILES changes. Use the list of MFILES found by find to identify the set of possible auto-generated files and add the intersection of this set and SRCS to CLEANFILES. Submitted by: imp (previous version), sbruno Differential Revision: https://reviews.freebsd.org/D4336
2015-12-01 21:19:16 +00:00
_MFILES=${__MPATH:T:O}
_MPATH=${__MPATH:H:O:u}
Kill MFILES and find things automatically. It turned out to be only lightly used. Find the proper .m file when we depend on *_if.[ch] in the srcs line, with seat-belts for false positive matches. This uses make's path mechanism. A further refinement would be to calculate this once, and then pass the resulting _MPATH to modules submakes. Differential Revision: https://reviews.freebsd.org/D2327
2015-07-03 01:50:26 +00:00
.PATH.m: ${_MPATH}
Fix cleaning of files generated from .m sources.
2015-08-28 13:57:30 +00:00
.for _i in ${SRCS:M*_if.[ch]}
Restore cleaning of auto-generated kobj files after the recent MFILES changes. Use the list of MFILES found by find to identify the set of possible auto-generated files and add the intersection of this set and SRCS to CLEANFILES. Submitted by: imp (previous version), sbruno Differential Revision: https://reviews.freebsd.org/D4336
2015-12-01 21:19:16 +00:00
_MATCH=M${_i:R:S/$/.m/}
_MATCHES=${_MFILES:${_MATCH}}
.if !empty(_MATCHES)
CLEANFILES+= ${_i}
.endif
Fix cleaning of files generated from .m sources.
2015-08-28 13:57:30 +00:00
.endfor # _i
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
.m.c: ${SYSDIR}/tools/makeobjops.awk
Kill MFILES and find things automatically. It turned out to be only lightly used. Find the proper .m file when we depend on *_if.[ch] in the srcs line, with seat-belts for false positive matches. This uses make's path mechanism. A further refinement would be to calculate this once, and then pass the resulting _MPATH to modules submakes. Differential Revision: https://reviews.freebsd.org/D2327
2015-07-03 01:50:26 +00:00
${AWK} -f ${SYSDIR}/tools/makeobjops.awk ${.IMPSRC} -c
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
.m.h: ${SYSDIR}/tools/makeobjops.awk
Kill MFILES and find things automatically. It turned out to be only lightly used. Find the proper .m file when we depend on *_if.[ch] in the srcs line, with seat-belts for false positive matches. This uses make's path mechanism. A further refinement would be to calculate this once, and then pass the resulting _MPATH to modules submakes. Differential Revision: https://reviews.freebsd.org/D2327
2015-07-03 01:50:26 +00:00
${AWK} -f ${SYSDIR}/tools/makeobjops.awk ${.IMPSRC} -h
Remove special case from *devs2h loop for usb Move usb out of the loop since the other devs2h programs don't grok -h Add new rule for usbdevs_data.h
2004-12-30 23:22:08 +00:00
.for _i in mii pccard
Use the more readable empty() syntax.
2005-09-07 17:38:46 +00:00
.if !empty(SRCS:M${_i}devs.h)
Move to generating pccarddevs.h on the fly, both for the kernel and the modules. Also generate usbdevs.h automatically now, but a non-kernel file is stopping that at the moment.
2004-05-26 00:53:10 +00:00
CLEANFILES+= ${_i}devs.h
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${_i}devs.h: ${SYSDIR}/tools/${_i}devs2h.awk ${SYSDIR}/dev/${_i}/${_i}devs
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/${_i}devs2h.awk ${SYSDIR}/dev/${_i}/${_i}devs
Remove miidevs.h and generate it from miidevs at compile time. The devlist2h.awk tool to do this has been repocopied to sys/tools/.
2003-01-19 02:59:34 +00:00
.endif
Move to generating pccarddevs.h on the fly, both for the kernel and the modules. Also generate usbdevs.h automatically now, but a non-kernel file is stopping that at the moment.
2004-05-26 00:53:10 +00:00
.endfor # _i
Remove miidevs.h and generate it from miidevs at compile time. The devlist2h.awk tool to do this has been repocopied to sys/tools/.
2003-01-19 02:59:34 +00:00
[bhnd] Initial bhnd(4) SPROM/NVRAM support. This adds support for the NVRAM handling and the basic SPROM hardware used on siba(4) and bcma(4) devices, including: * SPROM directly attached to the PCI core, accessible via PCI configuration space. * SPROM attached to later ChipCommon cores. * SPROM variables vended from the parent SoC bus (e.g. via a directly-attached flash device). Additional improvements to the NVRAM/SPROM interface will be required, but this changeset stands alone as working checkpoint. Submitted by: Landon Fuller <landonf@landonf.org> Reviewed by: Michael Zhilin <mizkha@gmail.com> (Broadcom MIPS support) Differential Revision: https://reviews.freebsd.org/D6196
2016-05-08 19:14:05 +00:00
.if !empty(SRCS:Mbhnd_nvram_map.h)
CLEANFILES+= bhnd_nvram_map.h
bhnd_nvram_map.h: ${SYSDIR}/dev/bhnd/tools/nvram_map_gen.awk \
${SYSDIR}/dev/bhnd/tools/nvram_map_gen.sh \
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${SYSDIR}/dev/bhnd/nvram/nvram_map
bhnd_nvram_map.h:
After r299241, which added bhnd(4), use sh to run the shell scripts for generating nvram maps, to allow a noexec-mounted source directory. Reported by: Johan Hendriks <joh.hendriks@gmail.com> PR: 209435
2016-05-13 16:35:37 +00:00
sh ${SYSDIR}/dev/bhnd/tools/nvram_map_gen.sh \
[bhnd] Initial bhnd(4) SPROM/NVRAM support. This adds support for the NVRAM handling and the basic SPROM hardware used on siba(4) and bcma(4) devices, including: * SPROM directly attached to the PCI core, accessible via PCI configuration space. * SPROM attached to later ChipCommon cores. * SPROM variables vended from the parent SoC bus (e.g. via a directly-attached flash device). Additional improvements to the NVRAM/SPROM interface will be required, but this changeset stands alone as working checkpoint. Submitted by: Landon Fuller <landonf@landonf.org> Reviewed by: Michael Zhilin <mizkha@gmail.com> (Broadcom MIPS support) Differential Revision: https://reviews.freebsd.org/D6196
2016-05-08 19:14:05 +00:00
${SYSDIR}/dev/bhnd/nvram/nvram_map -h
.endif
.if !empty(SRCS:Mbhnd_nvram_map_data.h)
CLEANFILES+= bhnd_nvram_map_data.h
bhnd_nvram_map_data.h: ${SYSDIR}/dev/bhnd/tools/nvram_map_gen.awk \
${SYSDIR}/dev/bhnd/tools/nvram_map_gen.sh \
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
${SYSDIR}/dev/bhnd/nvram/nvram_map
bhnd_nvram_map_data.h:
After r299241, which added bhnd(4), use sh to run the shell scripts for generating nvram maps, to allow a noexec-mounted source directory. Reported by: Johan Hendriks <joh.hendriks@gmail.com> PR: 209435
2016-05-13 16:35:37 +00:00
sh ${SYSDIR}/dev/bhnd/tools/nvram_map_gen.sh \
[bhnd] Initial bhnd(4) SPROM/NVRAM support. This adds support for the NVRAM handling and the basic SPROM hardware used on siba(4) and bcma(4) devices, including: * SPROM directly attached to the PCI core, accessible via PCI configuration space. * SPROM attached to later ChipCommon cores. * SPROM variables vended from the parent SoC bus (e.g. via a directly-attached flash device). Additional improvements to the NVRAM/SPROM interface will be required, but this changeset stands alone as working checkpoint. Submitted by: Landon Fuller <landonf@landonf.org> Reviewed by: Michael Zhilin <mizkha@gmail.com> (Broadcom MIPS support) Differential Revision: https://reviews.freebsd.org/D6196
2016-05-08 19:14:05 +00:00
${SYSDIR}/dev/bhnd/nvram/nvram_map -d
.endif
Use the more readable empty() syntax.
2005-09-07 17:38:46 +00:00
.if !empty(SRCS:Musbdevs.h)
Remove special case from *devs2h loop for usb Move usb out of the loop since the other devs2h programs don't grok -h Add new rule for usbdevs_data.h
2004-12-30 23:22:08 +00:00
CLEANFILES+= usbdevs.h
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
usbdevs.h: ${SYSDIR}/tools/usbdevs2h.awk ${SYSDIR}/dev/usb/usbdevs
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/usbdevs2h.awk ${SYSDIR}/dev/usb/usbdevs -h
Remove special case from *devs2h loop for usb Move usb out of the loop since the other devs2h programs don't grok -h Add new rule for usbdevs_data.h
2004-12-30 23:22:08 +00:00
.endif
Use the more readable empty() syntax.
2005-09-07 17:38:46 +00:00
.if !empty(SRCS:Musbdevs_data.h)
Remove special case from *devs2h loop for usb Move usb out of the loop since the other devs2h programs don't grok -h Add new rule for usbdevs_data.h
2004-12-30 23:22:08 +00:00
CLEANFILES+= usbdevs_data.h
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
usbdevs_data.h: ${SYSDIR}/tools/usbdevs2h.awk ${SYSDIR}/dev/usb/usbdevs
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/usbdevs2h.awk ${SYSDIR}/dev/usb/usbdevs -d
Remove special case from *devs2h loop for usb Move usb out of the loop since the other devs2h programs don't grok -h Add new rule for usbdevs_data.h
2004-12-30 23:22:08 +00:00
.endif
Use the more readable empty() syntax.
2005-09-07 17:38:46 +00:00
.if !empty(SRCS:Macpi_quirks.h)
Hook acpi_quirks up to the build for kernel and modules.
2004-06-24 06:29:22 +00:00
CLEANFILES+= acpi_quirks.h
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
acpi_quirks.h: ${SYSDIR}/tools/acpi_quirks2h.awk ${SYSDIR}/dev/acpica/acpi_quirks
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
${AWK} -f ${SYSDIR}/tools/acpi_quirks2h.awk ${SYSDIR}/dev/acpica/acpi_quirks
Hook acpi_quirks up to the build for kernel and modules.
2004-06-24 06:29:22 +00:00
.endif
Support assym.s in DPSRCS to depend on it but not link it in. Reported by: andrew MFC after: 2 weeks Sponsored by: Dell EMC Isilon
2017-04-05 17:56:17 +00:00
.if !empty(SRCS:Massym.s) || !empty(DPSRCS:Massym.s)
Add rules for building assym.s.
2005-04-22 09:45:39 +00:00
CLEANFILES+= assym.s genassym.o
kmod: Fix building assym.o not building missing dependencies. For instance, in the dtrace/dtrace module, building dtrace_asm.o wants to build genassym.o first, but it doesn't build the missing ilinks and if_*.h headers which are part of the OBJS_DEPEND_GUESS list of dependencies to build if a .depend file is missing. MFC after: 1 week Sponsored by: Dell EMC Isilon
2017-03-17 18:08:33 +00:00
DEPENDOBJS+= genassym.o
Add rules for building assym.s.
2005-04-22 09:45:39 +00:00
assym.s: genassym.o
Force an explicit dependency on opt_global.h for all module object files when building modules as part of a kernel build just as we do for kernel object files. MFC after: 1 week Reported by: kmacy, kris Reviewed by: ru
2008-03-04 16:54:31 +00:00
.if defined(KERNBUILDDIR)
genassym.o: opt_global.h
.endif
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
assym.s: ${SYSDIR}/kern/genassym.sh
Retire the '@' symlink. It isn't really needed and causes more problems than it solves. SYSDIR is already defined almost always and can be used instead. Working around the one case where it isn't is much easier than working around the fact that @ may not exist in 18 other places. Differential Revision: https://reviews.freebsd.org/D1100
2014-11-06 16:48:37 +00:00
sh ${SYSDIR}/kern/genassym.sh genassym.o > ${.TARGET}
Fix kernel build ${MACHINE} path $M should be the kernel machine src directory, ${MACHINE}. In most cases ${MACHINE} and ${MACHINE_CPUARCH} are the same, but this is not true for pc98 and arm64. It appears we previously set M=${MACHINE_CPUARCH} as a workaround to accommodate pc98, where MACHINE_CPUARCH is pc98 but it uses sys/i386/i386/genassym.c. arm64 relies on this being set correctly, so update $M and add explicit workarounds for pc98. Differential Revision: https://reviews.freebsd.org/D2307 Reviewed by: andrew, imp Sponsored by: The FreeBSD Foundation
2015-04-16 22:34:10 +00:00
genassym.o: ${SYSDIR}/${MACHINE}/${MACHINE}/genassym.c
Revert r301079. This breaks cross-building with WITH_META_MODE since it will rebuild 'build-tools' during the 'everything' phase. A more proper fix is coming to bmake to implicitly require .META unless .NOMETA (and other restrictions) are in place.
2016-06-03 19:25:36 +00:00
genassym.o: ${SRCS:Mopt_*.h}
Exclude -flto when building *genassym.o The build process generates *assym.h using nm from *genassym.o (which is in turn created from *genassym.c). When compiling with link-time optimization (LTO) using -flto, .o files are LLVM bitcode, not ELF objects. This is not usable by genassym.sh, so remove -flto from those ${CC} invocations. Submitted by: George Rimar Reviewed by: dim MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D9659
2017-02-21 18:59:17 +00:00
${CC} -c ${CFLAGS:N-flto:N-fno-common} \
Fix kernel build ${MACHINE} path $M should be the kernel machine src directory, ${MACHINE}. In most cases ${MACHINE} and ${MACHINE_CPUARCH} are the same, but this is not true for pc98 and arm64. It appears we previously set M=${MACHINE_CPUARCH} as a workaround to accommodate pc98, where MACHINE_CPUARCH is pc98 but it uses sys/i386/i386/genassym.c. arm64 relies on this being set correctly, so update $M and add explicit workarounds for pc98. Differential Revision: https://reviews.freebsd.org/D2307 Reviewed by: andrew, imp Sponsored by: The FreeBSD Foundation
2015-04-16 22:34:10 +00:00
${SYSDIR}/${MACHINE}/${MACHINE}/genassym.c
Add rules for building assym.s.
2005-04-22 09:45:39 +00:00
.endif
Allow "make lint" to mostly work. Our sources are very unclean WRT lint, so this is turned off by default. Setting WANT_LINT will turn on generation of lint libraries for /usr/libdata/lint/*.ln. Reviewd by: silence in -audit.
2002-03-17 10:05:57 +00:00
lint: ${SRCS}
lint the previous lint commit. Reviewed by: markm
2002-03-18 09:35:45 +00:00
${LINT} ${LINTKERNFLAGS} ${CFLAGS:M-[DILU]*} ${.ALLSRC:M*.c}
Allow "make lint" to mostly work. Our sources are very unclean WRT lint, so this is turned off by default. Setting WANT_LINT will turn on generation of lint libraries for /usr/libdata/lint/*.ln. Reviewd by: silence in -audit.
2002-03-17 10:05:57 +00:00
Force an explicit dependency on opt_global.h for all module object files when building modules as part of a kernel build just as we do for kernel object files. MFC after: 1 week Reported by: kmacy, kris Reviewed by: ru
2008-03-04 16:54:31 +00:00
.if defined(KERNBUILDDIR)
${OBJS}: opt_global.h
.endif
Remove ilinks in cleandepend directly via CLEANDEPENDFILES. The 'cleanilinks' target is kept since it may still be useful as added in r200178, though never documented. Sponsored by: EMC / Isilon Storage Division
2016-02-24 17:18:58 +00:00
CLEANDEPENDFILES+= ${_ILINKS}
- Remove include links only when .depend is also removed, so that "make depend; make clean; make -n" works. - Preseve kernel's .depend if it already exists and its creation is interrupted. Reported/reviewed by: bde
2006-10-13 22:28:14 +00:00
# .depend needs include links so we remove them only together.
cleanilinks:
rm -f ${_ILINKS}
FAST_DEPEND: Rework how guessed dependencies are handled. Rather than depend on .depend not existing, check the actual .depend.OBJ file that will be used for that object. If it doesn't exist then use the guessed dependencies. FAST_DEPEND may never have a .depend file. Not having one means all of the previous logic would over-depend all object files on all headers which is not what we wanted. It also means that if a .depend is generated before a build is done for _EXTRADEPEND (such as for PROG or LIB) then all of these dependencies would not be used since the .depend wasn't generated from mkdep and the real .depend.* files are not generated until the build. Sponsored by: EMC / Isilon Storage Division
2016-02-24 17:19:05 +00:00
OBJS_DEPEND_GUESS+= ${SRCS:M*.h}
Add guessed dependencies to OBJS after bsd.dep.mk in case of it adding to SRCS. This was a regression in r295985. bsd.dep.mk adds to SRCS for dtrace probes, yacc grammars and some others. The code that is moving is planned to be removed once FAST_DEPEND is default (and the only option) though since FAST_DEPEND doesn't use this. Pointyhat to: bdrewery Sponsored by: EMC / Isilon Storage Division
2016-03-02 21:03:42 +00:00
.include <bsd.dep.mk>
Import bsd.clang-analyze.mk based on NetBSD's version. This allows 'make analyze' or 'make OBJ.clang-analyzer' to run the Clang static analyzer and present results on stdout. Obtained from: NetBSD (CVS Rev. 1.3) Sponsored by: EMC / Isilon Storage Division Differential Revision: https://reviews.freebsd.org/D5449
2016-02-26 22:14:00 +00:00
.include <bsd.clang-analyze.mk>
Fixed `make depend', which I broke here about a month ago by moving the null default for `depend' from bsd.dep.mk to bsd.obj.mk. bsd.kmod.mk included these files in a bad order, so it picked up the null default.
1998-06-09 02:55:40 +00:00
.include <bsd.obj.mk>
Initiate the de-orbit burn sequence for <bsd.kern.mk>. Always use sys/conf/kern.mk when building kernel/modules. <bsd.kern.mk> is only preserved for sys/boot/pc98/boot2 for now, but this will be fixed. If there are other users of <bsd.kern.mk>, please let me know. Reminded by: bde
2003-02-28 22:12:17 +00:00
.include "kern.mk"
Reference in New Issue Copy Permalink