d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
0f33c9fb91
freebsd-dev/sys/conf/Makefile.i386

286 lines
7.3 KiB
Makefile
Raw Normal View History

Bring back the `config file in the kernel' feature from the 1.x days. This is conditionalized by the INCLUDE_CONFIG_FILE option in your kernel config file and is not turned on by default. Submitted-By: Bill Pechter <pechter@shell.monmouth.com>
1996-06-08 23:27:16 +00:00
# Makefile.i386 -- with config changes.
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
# Copyright 1990 W. Jolitz
This Makefile is for FreeBSD, not 4.3 BSD-Reno. Moved $Id$. Added STRIP=, DBSYM=, and LOAD_ADDRESS?= Now use LOAD_ADDRESS for linking kernel and for dbsym, added strip -x to cut kernel size. Added machde.o: dependency, this will be needed in the future, and for now it does not hurt anyone. Cleaned out conf.o: dependency, mkdep does the right things. Same for param.c: This is really a Merge in of NetBSD's Makefile.i386, here is the relevant rlog info: ---------------------------- revision 1.27 date: 1993/08/27 23:58:20; author: brezak; state: Exp; lines: +2 -2 Need LOAD_ADDRESS for depend pass. ---------------------------- revision 1.25 date: 1993/07/19 16:52:16; author: mycroft; state: Exp; lines: +3 -3 Add ${DEBUG} to CFLAGS and -f to dbsym. ---------------------------- revision 1.22 date: 1993/07/18 10:08:22; author: mycroft; state: Exp; lines: +5 -6 Change to work with new config stuff for specifying load address. ---------------------------- revision 1.20 date: 1993/07/18 09:47:40; author: mycroft; state: Exp; lines: +6 -5 Use new -T option to dbsym. ---------------------------- revision 1.17 date: 1993/07/11 08:42:22; author: cgd; state: Exp; lines: +2 -2 don't ignore errors from dbsym... it might say that, e.g. there's not enough symbol space! ---------------------------- revision 1.14 date: 1993/06/06 23:29:03; author: cgd; state: Exp; lines: +2 -2 make conf.o actually depend on conf.c... ---------------------------- revision 1.8 date: 1993/04/29 03:27:39; author: cgd; state: Exp; lines: +5 -10 use ed instead of ex. the script to use is identical, and we might want to switch back to using ex when our ex supports -. ---------------------------- revision 1.5 date: 1993/03/24 18:48:57; author: cgd; state: Exp; lines: +1 -1 now use absolute path for dbsym ----------------------------
1993-09-26 19:44:33 +00:00
# from: @(#)Makefile.i386 7.1 5/10/91
$Id$ -> $FreeBSD$
1999-08-28 01:08:13 +00:00
# $FreeBSD$
This Makefile is for FreeBSD, not 4.3 BSD-Reno. Moved $Id$. Added STRIP=, DBSYM=, and LOAD_ADDRESS?= Now use LOAD_ADDRESS for linking kernel and for dbsym, added strip -x to cut kernel size. Added machde.o: dependency, this will be needed in the future, and for now it does not hurt anyone. Cleaned out conf.o: dependency, mkdep does the right things. Same for param.c: This is really a Merge in of NetBSD's Makefile.i386, here is the relevant rlog info: ---------------------------- revision 1.27 date: 1993/08/27 23:58:20; author: brezak; state: Exp; lines: +2 -2 Need LOAD_ADDRESS for depend pass. ---------------------------- revision 1.25 date: 1993/07/19 16:52:16; author: mycroft; state: Exp; lines: +3 -3 Add ${DEBUG} to CFLAGS and -f to dbsym. ---------------------------- revision 1.22 date: 1993/07/18 10:08:22; author: mycroft; state: Exp; lines: +5 -6 Change to work with new config stuff for specifying load address. ---------------------------- revision 1.20 date: 1993/07/18 09:47:40; author: mycroft; state: Exp; lines: +6 -5 Use new -T option to dbsym. ---------------------------- revision 1.17 date: 1993/07/11 08:42:22; author: cgd; state: Exp; lines: +2 -2 don't ignore errors from dbsym... it might say that, e.g. there's not enough symbol space! ---------------------------- revision 1.14 date: 1993/06/06 23:29:03; author: cgd; state: Exp; lines: +2 -2 make conf.o actually depend on conf.c... ---------------------------- revision 1.8 date: 1993/04/29 03:27:39; author: cgd; state: Exp; lines: +5 -10 use ed instead of ex. the script to use is identical, and we might want to switch back to using ex when our ex supports -. ---------------------------- revision 1.5 date: 1993/03/24 18:48:57; author: cgd; state: Exp; lines: +1 -1 now use absolute path for dbsym ----------------------------
1993-09-26 19:44:33 +00:00
#
# Makefile for FreeBSD
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
#
# This makefile is constructed from a machine description:
# config machineid
# Most changes should be made in the machine description
# /sys/i386/conf/``machineid''
# after which you should do
# config machineid
# Generic makefile changes should be made in
# /sys/i386/conf/Makefile.i386
# after which config should be rerun for all machines.
#
Make the necessary changes so that an ELF kernel can be built. I have successfully built, booted, and run a number of different ELF kernel configurations, including GENERIC. LINT also builds and links cleanly, though I have not tried to boot it. The impact on developers is virtually nil, except for two things. All linker sets that might possibly be present in the kernel must be listed in "sys/i386/i386/setdefs.h". And all C symbols that are also referenced from assembly language code must be listed in "sys/i386/include/asnames.h". It so happens that failure to do these things will have no impact on the a.out kernel. But it will break the build of the ELF kernel. The ELF bootloader works, but it is not ready to commit quite yet.
1997-04-22 06:55:47 +00:00
Put in an initial %VERSREQ field. This will cause existing config(8)'s to give a non-fatal warning about unknown directives, so there is no hurry to rebuild config(8) yet, apart from shutting up the warning.
1997-10-22 00:48:36 +00:00
# Which version of config(8) is required.
Move a couple of bits of (now) static content from config(8) (ie: %LOAD) directly into the Makefile. Remove references to swapkernel.c, it's not generated by config(8) now. (The previous config commits had generated it, but they had an unused 'char *' in them).
1999-05-09 18:57:02 +00:00
%VERSREQ= 400016
Put in an initial %VERSREQ field. This will cause existing config(8)'s to give a non-fatal warning about unknown directives, so there is no hurry to rebuild config(8) yet, apart from shutting up the warning.
1997-10-22 00:48:36 +00:00
Move KERNEL?=kernel to top of the file where it's more obvious and add a pointer to 'makeoptions' and /etc/make.conf. Catch a few stray "kernel" hardcoded references. Move the kernel.debug and related kernel build rules together.
1999-05-09 22:42:17 +00:00
# Can be overridden by makeoptions or /etc/make.conf
KERNEL?= kernel
Flip the kernel default to ELF.. Add a test to try and warn people that they might be about to blow their feet off if they have not been reading their mail. I don't know if or how well this will work, but it's worth a try.
1999-01-07 06:52:41 +00:00
KERNFORMAT?= elf
Add option for compiling in a 8x16 font.
1997-07-25 11:53:30 +00:00
STD8X16FONT?= iso
Allow the source root `S' to be overridden by defining it only when it isn't already defined. It enables config(8) to create the kernel build directory where it wants.
1999-10-30 09:28:27 +00:00
.if !defined(S)
Change all #includes to follow the current Berkeley style. Some of these ``changes'' are actually not changes at all, but CVS sometimes has trouble telling the difference. This also includes support for second-directory compiles. This is not quite complete yet, as `config' doesn't yet do the right thing. You can still make it work trivially, however, by doing the following: rm /sys/compile mkdir /usr/obj/sys/compile ln -s M-. /sys/compile cd /sys/i386/conf config MYKERNEL cd ../../compile/MYKERNEL ln -s /sys @ rm machine ln -s @/i386/include machine make depend make
1994-08-13 03:50:34 +00:00
.if exists(./@/.)
S= ./@
.else
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
S= ../..
Change all #includes to follow the current Berkeley style. Some of these ``changes'' are actually not changes at all, but CVS sometimes has trouble telling the difference. This also includes support for second-directory compiles. This is not quite complete yet, as `config' doesn't yet do the right thing. You can still make it work trivially, however, by doing the following: rm /sys/compile mkdir /usr/obj/sys/compile ln -s M-. /sys/compile cd /sys/i386/conf config MYKERNEL cd ../../compile/MYKERNEL ln -s /sys @ rm machine ln -s @/i386/include machine make depend make
1994-08-13 03:50:34 +00:00
.endif
Allow the source root `S' to be overridden by defining it only when it isn't already defined. It enables config(8) to create the kernel build directory where it wants.
1999-10-30 09:28:27 +00:00
.endif
Change all #includes to follow the current Berkeley style. Some of these ``changes'' are actually not changes at all, but CVS sometimes has trouble telling the difference. This also includes support for second-directory compiles. This is not quite complete yet, as `config' doesn't yet do the right thing. You can still make it work trivially, however, by doing the following: rm /sys/compile mkdir /usr/obj/sys/compile ln -s M-. /sys/compile cd /sys/i386/conf config MYKERNEL cd ../../compile/MYKERNEL ln -s /sys @ rm machine ln -s @/i386/include machine make depend make
1994-08-13 03:50:34 +00:00
I386= ${S}/i386
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
SIZE?= size
Go back to -O optimization...Bruce's arguments against -O2 are convincing enough for me.
1994-11-27 14:26:16 +00:00
COPTFLAGS?=-O
Added -I- to ${INCLUDES}. This makes 4.4Lite's (mis?)use of "foo.h" for headers in the compile directory work unsurprisingly. Without -I-, the search for "foo.h" begins in the directory of the file that includes it, and the compile directory is only searched because `-I.' is in ${INCLUDES}. Removed -I$S/sys from ${INCLUDES}. It was once necessary to find things like "param.h" in $S/sys. Now <sys/param.h> is found in $S.
1996-09-27 16:34:16 +00:00
INCLUDES= -nostdinc -I- -I. -I$S
Modify behavior of INCLUDES to never look in /usr/include unless we can't find the src/include directory. Reviewed by: "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com> Submitted by: Mike Pritchard <pritc003@maroon.tc.umn.edu>
1995-04-05 04:10:58 +00:00
# This hack is to allow kernel compiles to succeed on machines w/out srcdist
.if exists($S/../include)
INCLUDES+= -I$S/../include
.else
INCLUDES+= -I/usr/include
.endif
KAME netinet6 basic part(no IPsec,no V6 Multicast Forwarding, no UDP/TCP for IPv6 yet) With this patch, you can assigne IPv6 addr automatically, and can reply to IPv6 ping. Reviewed by: freebsd-arch, cvs-committers Obtained from: KAME project
1999-11-22 02:45:11 +00:00
COPTS= ${INCLUDES} ${IDENT} -DKERNEL -include opt_global.h
# KAME mandatory flags
COPTS+= -D_KERNEL
Added CWARNFLAGS and COPTFLAGS defines to Makefile so that we can easily test various warning and optimization compiler flags without modifying the Makefile. (From my patch set.)
1993-11-07 04:41:11 +00:00
CFLAGS= ${COPTFLAGS} ${CWARNFLAGS} ${DEBUG} ${COPTS}
Change BINFORMAT to KERNFORMAT and add -aout or -elf to the tool arguments as appropriate. This is necessary because setting OBJFORMAT=aout in the makefile doesn't get exported to the environment. I wish our make could do that!
1998-09-01 00:11:02 +00:00
Fixed missing dependencies of genassym.o, gensetdefs.o and vers.o on opt_global.h. This actually matters for genassym.o (it depends on at least SMP). Don't undefine KERNEL for compiling genassym.c. genassym.o really depends on KERNEL, and the prototype mismatches that required undefining KERNEL in rev.1.49 no longer exist. Compile gensetdefs* with the same flags as genassym*. External `gen' programs such linux_genassym should also use these flags (${GEN_CFLAGS}). Fixed missing dependency generation for gensetdefs.o. C sources for external `gen' programs should be added to GEN_CFILES to get their dependencies generated. Cleaned up flags definitions and use. All of ${CFLAGS} is now passed to mkdep and the assembler (both are actually variants of ${CC} and will ignore the irrelevant flags).
1998-09-21 06:39:11 +00:00
# XXX LOCORE means "don't declare C stuff" not "for locore.s".
ASM_CFLAGS= -x assembler-with-cpp -DLOCORE ${CFLAGS}
# Use the default object format for genassym, etc.
Use the same -UKERNEL strategy as the alpha to avoid the inlines etc.
1999-06-28 09:21:41 +00:00
GEN_CFLAGS= ${COPTFLAGS} ${CWARNFLAGS} ${DEBUG} ${COPTS} -UKERNEL
Decouple genassym flags from CFLAGS. Submitted by: jhay
1998-09-14 11:32:17 +00:00
Change BINFORMAT to KERNFORMAT and add -aout or -elf to the tool arguments as appropriate. This is necessary because setting OBJFORMAT=aout in the makefile doesn't get exported to the environment. I wish our make could do that!
1998-09-01 00:11:02 +00:00
# Select the correct set of tools. Can't set OBJFORMAT here because it
Fixed missing dependencies of genassym.o, gensetdefs.o and vers.o on opt_global.h. This actually matters for genassym.o (it depends on at least SMP). Don't undefine KERNEL for compiling genassym.c. genassym.o really depends on KERNEL, and the prototype mismatches that required undefining KERNEL in rev.1.49 no longer exist. Compile gensetdefs* with the same flags as genassym*. External `gen' programs such linux_genassym should also use these flags (${GEN_CFLAGS}). Fixed missing dependency generation for gensetdefs.o. C sources for external `gen' programs should be added to GEN_CFILES to get their dependencies generated. Cleaned up flags definitions and use. All of ${CFLAGS} is now passed to mkdep and the assembler (both are actually variants of ${CC} and will ignore the irrelevant flags).
1998-09-21 06:39:11 +00:00
# doesn't get exported into the environment, and if it were exported
# then it might break building of genassym, etc.
Change BINFORMAT to KERNFORMAT and add -aout or -elf to the tool arguments as appropriate. This is necessary because setting OBJFORMAT=aout in the makefile doesn't get exported to the environment. I wish our make could do that!
1998-09-01 00:11:02 +00:00
.if ${KERNFORMAT} == "elf"
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
FMT= -elf
Change BINFORMAT to KERNFORMAT and add -aout or -elf to the tool arguments as appropriate. This is necessary because setting OBJFORMAT=aout in the makefile doesn't get exported to the environment. I wish our make could do that!
1998-09-01 00:11:02 +00:00
.else
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
FMT= -aout
Change BINFORMAT to KERNFORMAT and add -aout or -elf to the tool arguments as appropriate. This is necessary because setting OBJFORMAT=aout in the makefile doesn't get exported to the environment. I wish our make could do that!
1998-09-01 00:11:02 +00:00
.endif
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
CFLAGS+= ${FMT}
Change BINFORMAT to KERNFORMAT and add -aout or -elf to the tool arguments as appropriate. This is necessary because setting OBJFORMAT=aout in the makefile doesn't get exported to the environment. I wish our make could do that!
1998-09-01 00:11:02 +00:00
Increased kernel virtual address space to 1GB. NOTE: You MUST have fixed bootblocks in order to boot the kernel after this! Also note that this change breaks BSDI BSD/OS compatibility. Also increased default NKPT to 17 so that FreeBSD can boot on machines with >=2GB of RAM. Booting on machines with exactly 4GB requires other patches, not included.
1999-03-11 18:28:46 +00:00
LOAD_ADDRESS?= C0100000
Fixed the fix for not using -fomit-frame-pointer with -pg. The previous fix stopped it being used in all cases, because substitution on unset variables does not work. When profiling, put -malign-functions=4 in CFLAGS instead of in PROF. This fixes the histogram counts for profiling support functions. It gives bogus but harmless extra alignment for genassym etc.
1997-06-29 16:39:11 +00:00
DEFINED_PROF= ${PROF}
Moved nonstandard compiler profiling options out of config. Just print the profiling level in config and decide what to do in makefiles. Makefile.i386: Align functions to 16-byte boundaries if profiling is enabled. This will allow a fourfold reduction in the size of the profiling buffers.
1996-12-13 12:46:28 +00:00
.if defined(PROF)
Fixed the fix for not using -fomit-frame-pointer with -pg. The previous fix stopped it being used in all cases, because substitution on unset variables does not work. When profiling, put -malign-functions=4 in CFLAGS instead of in PROF. This fixes the histogram counts for profiling support functions. It gives bogus but harmless extra alignment for genassym etc.
1997-06-29 16:39:11 +00:00
CFLAGS+= -malign-functions=4
Moved nonstandard compiler profiling options out of config. Just print the profiling level in config and decide what to do in makefiles. Makefile.i386: Align functions to 16-byte boundaries if profiling is enabled. This will allow a fourfold reduction in the size of the profiling buffers.
1996-12-13 12:46:28 +00:00
.if ${PROFLEVEL} >= 2
IDENT+= -DGPROF4 -DGUPROF
PROF+= -mprofiler-epilogue
.endif
.endif
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Fixed "misspelling" of bcmp as memcmp. memcmp doesn't exist in the kernel, but gcc provides a pessimal builtin for it. Makefile.i386: Added a variable (CONF_CFLAGS) for configuration-specific compiler flags. LINT: Use CONF_CFLAGS to inhibit use of gcc builtins.
1999-10-03 07:09:31 +00:00
# Put configuration-specific C flags last (except for ${PROF}) so that they
# can override the others.
CFLAGS+= ${CONF_CFLAGS}
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
NORMAL_C= ${CC} -c ${CFLAGS} ${PROF} $<
Removed all traces of PARAM in Makefile.i386. Incremented CONFIGVERS to reflect the dependency of Makefile.i386 on nothing being put in PARAM. Config versioning is too closely coupled with the Makefile.i386.
1998-07-12 09:52:45 +00:00
NORMAL_C_C= ${CC} -c ${CFLAGS} ${PROF} $<
Fixed missing dependencies of genassym.o, gensetdefs.o and vers.o on opt_global.h. This actually matters for genassym.o (it depends on at least SMP). Don't undefine KERNEL for compiling genassym.c. genassym.o really depends on KERNEL, and the prototype mismatches that required undefining KERNEL in rev.1.49 no longer exist. Compile gensetdefs* with the same flags as genassym*. External `gen' programs such linux_genassym should also use these flags (${GEN_CFLAGS}). Fixed missing dependency generation for gensetdefs.o. C sources for external `gen' programs should be added to GEN_CFILES to get their dependencies generated. Cleaned up flags definitions and use. All of ${CFLAGS} is now passed to mkdep and the assembler (both are actually variants of ${CC} and will ignore the irrelevant flags).
1998-09-21 06:39:11 +00:00
NORMAL_S= ${CC} -c ${ASM_CFLAGS} $<
Removed all traces of PARAM in Makefile.i386. Incremented CONFIGVERS to reflect the dependency of Makefile.i386 on nothing being put in PARAM. Config versioning is too closely coupled with the Makefile.i386.
1998-07-12 09:52:45 +00:00
PROFILE_C= ${CC} -c ${CFLAGS} $<
Remove exception.o, swtch.o, support.o, conf.o and machdep.o from SYSTEM_OBS. They are now normal objects. Remove stale dependencies for the above now-normal objects and for locore.o and generate dependencies using mkdep. Config doesn't generate lists of assembler source files so the lists to be mkdep'ed have to be given explictly. Only the standard *.s files are given, so the dependencies for gnu/fpemul/*.s are incomplete. *.S files would be handled right if config put them in CFILES. Don't define NPX. It was replaced by NNPX > 0 years ago. Define LOAD_ADDRESS in COPTS so that compiling machdep.c isn't a special case. Moving around the dependencies exposed a bug in make. It doesn't know that assym.s and ./assym.s are the same. Add a rule tell it.
1994-08-30 17:17:49 +00:00
Use the gensetdefs that is now in /usr/bin, it's a superset of the one in the kernel source, and that one is already used for modules. I don't _think_ this will hurt releases, aout-to-elf, etc, but it is possible. In all the cases I've looked at, config(8) has been generated straight after a make world, so if /usr/sbin/config exists and is the right version for the kernel, then we can pretty much count on /usr/bin/gensetdefs being there too.
1999-01-17 21:17:27 +00:00
GEN_CFILES= ${I386}/i386/genassym.c
# setdef0.c and setdef1.c are intentionally
Fixed missing dependencies of genassym.o, gensetdefs.o and vers.o on opt_global.h. This actually matters for genassym.o (it depends on at least SMP). Don't undefine KERNEL for compiling genassym.c. genassym.o really depends on KERNEL, and the prototype mismatches that required undefining KERNEL in rev.1.49 no longer exist. Compile gensetdefs* with the same flags as genassym*. External `gen' programs such linux_genassym should also use these flags (${GEN_CFLAGS}). Fixed missing dependency generation for gensetdefs.o. C sources for external `gen' programs should be added to GEN_CFILES to get their dependencies generated. Cleaned up flags definitions and use. All of ${CFLAGS} is now passed to mkdep and the assembler (both are actually variants of ${CC} and will ignore the irrelevant flags).
1998-09-21 06:39:11 +00:00
# omitted from SYSTEM_CFILES. They include setdefs.h, a header which
This commit affects ELF kernels only. Remove "setdefs.h" and arrange to generate it automatically at ELF kernel build time. "gensetdefs.c" is a utility which scans a set of ELF object files and outputs a line ``DEFINE_SET(name, length);'' for each linker set that it finds. When generating an ELF kernel, this is run just before the final link to generate "setdefs.h". Remove the init_sets() function from "setdef0.c", and its call from "machdep.c". Since "gensetdefs.c" calculates the length of each set, it is no longer necessary in an ELF kernel to count the set elements at kernel initialization time. Also remove "set_of_sets" which was used for this purpose. Link "setdef0" and "setdef1" into the kernel only if building for ELF. Since init_sets() is no longer used, there is no need to link them into an a.out kernel.
1997-05-21 23:21:30 +00:00
# is generated from all of ${OBJS}. We don't want to have to compile
# everything just to do a make depend.
SYSTEM_CFILES= ioconf.c param.c vnode_if.c config.c
Remove exception.o, swtch.o, support.o, conf.o and machdep.o from SYSTEM_OBS. They are now normal objects. Remove stale dependencies for the above now-normal objects and for locore.o and generate dependencies using mkdep. Config doesn't generate lists of assembler source files so the lists to be mkdep'ed have to be given explictly. Only the standard *.s files are given, so the dependencies for gnu/fpemul/*.s are incomplete. *.S files would be handled right if config put them in CFILES. Don't define NPX. It was replaced by NNPX > 0 years ago. Define LOAD_ADDRESS in COPTS so that compiling machdep.c isn't a special case. Moving around the dependencies exposed a bug in make. It doesn't know that assym.s and ./assym.s are the same. Add a rule tell it.
1994-08-30 17:17:49 +00:00
SYSTEM_SFILES= ${I386}/i386/locore.s
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
SYSTEM_DEP= Makefile ${SYSTEM_OBJS}
Clean up the kernel/kernel.debug/install.debug etc stuff. Sense ${DEBUG} to decide on building kernel.debug or not. Use a common install and install.debug target to minimize duplication. (I deleted the ELF transition kernel warning, it was getting dated..)
1999-04-13 18:25:08 +00:00
Make a.out kernels kld compatable by default...
1998-11-04 15:27:41 +00:00
.if ${KERNFORMAT} == aout || ${KERNFORMAT} == aoutkld
This commit affects ELF kernels only. Remove "setdefs.h" and arrange to generate it automatically at ELF kernel build time. "gensetdefs.c" is a utility which scans a set of ELF object files and outputs a line ``DEFINE_SET(name, length);'' for each linker set that it finds. When generating an ELF kernel, this is run just before the final link to generate "setdefs.h". Remove the init_sets() function from "setdef0.c", and its call from "machdep.c". Since "gensetdefs.c" calculates the length of each set, it is no longer necessary in an ELF kernel to count the set elements at kernel initialization time. Also remove "set_of_sets" which was used for this purpose. Link "setdef0" and "setdef1" into the kernel only if building for ELF. Since init_sets() is no longer used, there is no need to link them into an a.out kernel.
1997-05-21 23:21:30 +00:00
SYSTEM_OBJS= locore.o vnode_if.o ${OBJS} ioconf.o param.o config.o
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
SYSTEM_LD= @${LD} ${FMT} -Bforcedynamic -Z -T ${LOAD_ADDRESS} -o ${.TARGET} -X ${SYSTEM_OBJS} vers.o
SYSTEM_LD_TAIL= @${SIZE} ${FMT} ${.TARGET} ; chmod 755 ${.TARGET}
This is the kernel linker. To use it, you will first need to apply the patches in freefall:/home/dfr/ld.diffs to your ld sources and set BINFORMAT to aoutkld when linking the kernel. Library changes and userland utilities will appear in a later commit.
1997-05-07 16:05:47 +00:00
.endif
Clean up the kernel/kernel.debug/install.debug etc stuff. Sense ${DEBUG} to decide on building kernel.debug or not. Use a common install and install.debug target to minimize duplication. (I deleted the ELF transition kernel warning, it was getting dated..)
1999-04-13 18:25:08 +00:00
Change BINFORMAT to KERNFORMAT and add -aout or -elf to the tool arguments as appropriate. This is necessary because setting OBJFORMAT=aout in the makefile doesn't get exported to the environment. I wish our make could do that!
1998-09-01 00:11:02 +00:00
.if ${KERNFORMAT} == elf
Regrettably, you can't take setdef0.c out of order, or, surprise, linker sets don't work and thus neither does the kernel. This should stop being a problem with an improved linker set mechanism in the pipeline (see the bootloader) but for now this has to stay like this.
1998-12-21 10:29:09 +00:00
SYSTEM_OBJS= locore.o setdef0.o vnode_if.o ${OBJS} ioconf.o param.o config.o \
Switch to using ".So" as the extension for PIC object files rather than ".so". The old extension conflicted with well-established naming conventions for dynamically loadable modules. The "clean" targets continue to remove ".so" files too, to deal with old systems.
1999-01-09 21:51:00 +00:00
setdef1.o hack.So
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
SYSTEM_LD= @${LD} ${FMT} -Bdynamic -T $S/i386/conf/kernel.script \
Make the ELF kernel build produce a dynamic executable (!). This enables the in-kernel linker to access the _DYNAMIC data for doing loadable elf modules. The alpha kernel is already done this way, I've borrowed some of the hacks from there. This is primarily aimed at the 3-stage boot process which is intended to be able to do pre-loading of kernel modules. Note that the entry point isn't 0xf0100000 any more, it'll be a little further on - but this value is stored in the headers. I don't think this will be a problem, but I'm sure somebody will tell me if it is. :-) I'm not sure if btxboot is going to like this, it doesn't do proper ELF header checking and assumes that there are exactly two program header entries and that they are both PT_LOAD entries - a bad assumption.
1998-09-30 12:14:39 +00:00
-export-dynamic -dynamic-linker /red/herring \
Use ${.TARGET} rather than $@. i386 tested, pc98 untested because config on my box doesn't grok machine type pc98. PR: 3272 Submitted by: jhs
1998-04-29 18:21:35 +00:00
-o ${.TARGET} -X ${SYSTEM_OBJS} vers.o
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
SYSTEM_LD_TAIL= @${SIZE} ${FMT} ${.TARGET} ; chmod 755 ${.TARGET}
Make the ELF kernel build produce a dynamic executable (!). This enables the in-kernel linker to access the _DYNAMIC data for doing loadable elf modules. The alpha kernel is already done this way, I've borrowed some of the hacks from there. This is primarily aimed at the 3-stage boot process which is intended to be able to do pre-loading of kernel modules. Note that the entry point isn't 0xf0100000 any more, it'll be a little further on - but this value is stored in the headers. I don't think this will be a problem, but I'm sure somebody will tell me if it is. :-) I'm not sure if btxboot is going to like this, it doesn't do proper ELF header checking and assumes that there are exactly two program header entries and that they are both PT_LOAD entries - a bad assumption.
1998-09-30 12:14:39 +00:00
SYSTEM_DEP+= $S/i386/conf/kernel.script
Make the necessary changes so that an ELF kernel can be built. I have successfully built, booted, and run a number of different ELF kernel configurations, including GENERIC. LINT also builds and links cleanly, though I have not tried to boot it. The impact on developers is virtually nil, except for two things. All linker sets that might possibly be present in the kernel must be listed in "sys/i386/i386/setdefs.h". And all C symbols that are also referenced from assembly language code must be listed in "sys/i386/include/asnames.h". It so happens that failure to do these things will have no impact on the a.out kernel. But it will break the build of the ELF kernel. The ELF bootloader works, but it is not ready to commit quite yet.
1997-04-22 06:55:47 +00:00
.endif
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Handle the aic7770 driver's dependancies correctly. YOU MUST REBUILD CONFIG.
1994-12-31 19:27:29 +00:00
%BEFORE_DEPEND
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
%OBJS
%CFILES
Use the new config(8) %SFILES token to auto-generate the SFILES list. Note: THIS REQUIRES THAT YOU REBUILD CONFIG(8)!!!
1995-10-29 11:10:53 +00:00
%SFILES
Reviewed by: Doug Rabson Submitted by: nsouch Adding I2C and SMB entries to LINT, CONFIGVERS modified in Makefile.i386
1998-09-03 20:58:34 +00:00
%MFILES
Fix "dependency" spelling error. Implement "clean" entries for device config entries. Add clean rules to aic7xxx and aic7xxx_seq.h. Submitted by: Pointed out by Bruce Evans <bde@zeta.org.au>
1995-03-01 10:08:07 +00:00
%CLEAN
Move a couple of bits of (now) static content from config(8) (ie: %LOAD) directly into the Makefile. Remove references to swapkernel.c, it's not generated by config(8) now. (The previous config commits had generated it, but they had an unused 'char *' in them).
1999-05-09 18:57:02 +00:00
all: ${KERNEL}
Move KERNEL?=kernel to top of the file where it's more obvious and add a pointer to 'makeoptions' and /etc/make.conf. Catch a few stray "kernel" hardcoded references. Move the kernel.debug and related kernel build rules together.
1999-05-09 22:42:17 +00:00
.if !defined(DEBUG)
FULLKERNEL= ${KERNEL}
.else
FULLKERNEL= ${KERNEL}.debug
${KERNEL}: ${FULLKERNEL}
.if ${KERNFORMAT} == "elf"
objcopy --strip-debug ${FULLKERNEL} ${KERNEL}
.else
cp ${FULLKERNEL} ${KERNEL}
strip -d ${KERNEL}
.endif
.endif
Hmm, might as well make all of BEFORE_DEPEND first in case make depend wasn't run.
1999-06-28 07:19:51 +00:00
${FULLKERNEL}: ${BEFORE_DEPEND} ${SYSTEM_DEP} vers.o
Move a couple of bits of (now) static content from config(8) (ie: %LOAD) directly into the Makefile. Remove references to swapkernel.c, it's not generated by config(8) now. (The previous config commits had generated it, but they had an unused 'char *' in them).
1999-05-09 18:57:02 +00:00
@rm -f ${.TARGET}
@echo linking ${.TARGET}
${SYSTEM_LD}
${SYSTEM_LD_TAIL}
If there is no .depends file, use the standard ad-hoc way of ensuring that the generated files are generated before any of the object files. Also minor cleanup of dependencies in conf/files that I bogusly added before. This should fix the requirement that make depend be done starting from a clean config directory. If you don't have a clean directory, make depend is still required if you want the proper .o's to be recompiled. Reviewed by: bde
1998-12-07 22:36:54 +00:00
.if !exists(.depend)
Ensure that vnode_if.h build before object files, if make depend has not been run.
1999-03-01 09:54:51 +00:00
${SYSTEM_OBJS}: vnode_if.h ${BEFORE_DEPEND:M*.h}
If there is no .depends file, use the standard ad-hoc way of ensuring that the generated files are generated before any of the object files. Also minor cleanup of dependencies in conf/files that I bogusly added before. This should fix the requirement that make depend be done starting from a clean config directory. If you don't have a clean directory, make depend is still required if you want the proper .o's to be recompiled. Reviewed by: bde
1998-12-07 22:36:54 +00:00
.endif
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
clean:
Switch to using ".So" as the extension for PIC object files rather than ".so". The old extension conflicted with well-established naming conventions for dynamically loadable modules. The "clean" targets continue to remove ".so" files too, to deal with old systems.
1999-01-09 21:51:00 +00:00
rm -f *.o *.so *.So *.ko *.s eddep errs genassym gensetdefs \
1. Modify config to issue different code for debugging. 2. Config complains if you use -g: Debugging is enabled by default, there is no ned to specify the -g option 3. Config warns you if you don't use -s: Building kernel with full debugging symbols. Do "config -s BSD" for historic partial symbolic support. To install the debugging kernel, do make install.debug (BSD was the name of the config file I used; I print out the same name). 4. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 and config to work if a kernel name other than 'kernel' is specified. This is not absolutely necessary, but useful, and it was relatively easy. I now have a kernel called /crapshit :-) 5. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 "clean" target to remove both the debug and normal kernel. 6. Modify all to install the stripped kernel by default and the debug kernel if you enter "make install.debug". 7. Update version number of Makefiles and config.
1999-04-07 09:28:03 +00:00
${KERNEL} ${FULLKERNEL} linterrs makelinks param.c \
setdef[01].c setdefs.h symbols.exclude symbols.sort tags \
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
vers.c vnode_if.c vnode_if.h ${CLEAN}
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Try to be a little smarter about touching the date on assym.s, so that if something changes which doesn't affect it, locore doesn't have to get rebuilt. This is at the cost of a genassym and a cmp in every compile, until someone can figure out how to make `make' smarter itself.
1993-11-07 16:46:33 +00:00
#lint: /tmp param.c
Removed all traces of PARAM in Makefile.i386. Incremented CONFIGVERS to reflect the dependency of Makefile.i386 on nothing being put in PARAM. Config versioning is too closely coupled with the Makefile.i386.
1998-07-12 09:52:45 +00:00
# @lint -hbxn -DGENERIC -Dvolatile= ${COPTS} \
Try to be a little smarter about touching the date on assym.s, so that if something changes which doesn't affect it, locore doesn't have to get rebuilt. This is at the cost of a genassym and a cmp in every compile, until someone can figure out how to make `make' smarter itself.
1993-11-07 16:46:33 +00:00
# ${I386}/i386/Locore.c ${CFILES} ioconf.c param.c | \
# grep -v 'struct/union .* never defined' | \
# grep -v 'possible pointer alignment problem'
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Restore some explicit dependencies for locore.s so that "make clean; make" works even if .depend doesn't exist.
1994-09-02 20:24:15 +00:00
locore.o: ${I386}/i386/locore.s assym.s
Remove exception.o, swtch.o, support.o, conf.o and machdep.o from SYSTEM_OBS. They are now normal objects. Remove stale dependencies for the above now-normal objects and for locore.o and generate dependencies using mkdep. Config doesn't generate lists of assembler source files so the lists to be mkdep'ed have to be given explictly. Only the standard *.s files are given, so the dependencies for gnu/fpemul/*.s are incomplete. *.S files would be handled right if config put them in CFILES. Don't define NPX. It was replaced by NNPX > 0 years ago. Define LOAD_ADDRESS in COPTS so that compiling machdep.c isn't a special case. Moving around the dependencies exposed a bug in make. It doesn't know that assym.s and ./assym.s are the same. Add a rule tell it.
1994-08-30 17:17:49 +00:00
${NORMAL_S}
Make the ELF kernel build produce a dynamic executable (!). This enables the in-kernel linker to access the _DYNAMIC data for doing loadable elf modules. The alpha kernel is already done this way, I've borrowed some of the hacks from there. This is primarily aimed at the 3-stage boot process which is intended to be able to do pre-loading of kernel modules. Note that the entry point isn't 0xf0100000 any more, it'll be a little further on - but this value is stored in the headers. I don't think this will be a problem, but I'm sure somebody will tell me if it is. :-) I'm not sure if btxboot is going to like this, it doesn't do proper ELF header checking and assumes that there are exactly two program header entries and that they are both PT_LOAD entries - a bad assumption.
1998-09-30 12:14:39 +00:00
.if ${KERNFORMAT} == elf
# This is a hack. BFD "optimizes" away dynamic mode if there are no
# dynamic references. We could probably do a '-Bforcedynamic' mode like
# in the a.out ld. For now, this works.
Switch to using ".So" as the extension for PIC object files rather than ".so". The old extension conflicted with well-established naming conventions for dynamically loadable modules. The "clean" targets continue to remove ".so" files too, to deal with old systems.
1999-01-09 21:51:00 +00:00
hack.So: Makefile
Make the ELF kernel build produce a dynamic executable (!). This enables the in-kernel linker to access the _DYNAMIC data for doing loadable elf modules. The alpha kernel is already done this way, I've borrowed some of the hacks from there. This is primarily aimed at the 3-stage boot process which is intended to be able to do pre-loading of kernel modules. Note that the entry point isn't 0xf0100000 any more, it'll be a little further on - but this value is stored in the headers. I don't think this will be a problem, but I'm sure somebody will tell me if it is. :-) I'm not sure if btxboot is going to like this, it doesn't do proper ELF header checking and assumes that there are exactly two program header entries and that they are both PT_LOAD entries - a bad assumption.
1998-09-30 12:14:39 +00:00
touch hack.c
Update the Alpha makefile to be more in line with the x86 version. It was getting quite dated. Attempt to minimize diffs between the two so it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
${CC} ${FMT} -shared -nostdlib hack.c -o hack.So
Make the ELF kernel build produce a dynamic executable (!). This enables the in-kernel linker to access the _DYNAMIC data for doing loadable elf modules. The alpha kernel is already done this way, I've borrowed some of the hacks from there. This is primarily aimed at the 3-stage boot process which is intended to be able to do pre-loading of kernel modules. Note that the entry point isn't 0xf0100000 any more, it'll be a little further on - but this value is stored in the headers. I don't think this will be a problem, but I'm sure somebody will tell me if it is. :-) I'm not sure if btxboot is going to like this, it doesn't do proper ELF header checking and assumes that there are exactly two program header entries and that they are both PT_LOAD entries - a bad assumption.
1998-09-30 12:14:39 +00:00
rm -f hack.c
.endif
Force the order of the setdefs* so that make -jN doesn't build the setdefs* multiple times at the same time that it's compiling setdefs[0].c. I think this was leading to unterminated linker sets etc.
1999-01-25 04:08:28 +00:00
.ORDER: setdefs.h setdef0.c setdef1.c
Use the gensetdefs that is now in /usr/bin, it's a superset of the one in the kernel source, and that one is already used for modules. I don't _think_ this will hurt releases, aout-to-elf, etc, but it is possible. In all the cases I've looked at, config(8) has been generated straight after a make world, so if /usr/sbin/config exists and is the right version for the kernel, then we can pretty much count on /usr/bin/gensetdefs being there too.
1999-01-17 21:17:27 +00:00
setdef0.o: setdef0.c setdefs.h
Make the necessary changes so that an ELF kernel can be built. I have successfully built, booted, and run a number of different ELF kernel configurations, including GENERIC. LINT also builds and links cleanly, though I have not tried to boot it. The impact on developers is virtually nil, except for two things. All linker sets that might possibly be present in the kernel must be listed in "sys/i386/i386/setdefs.h". And all C symbols that are also referenced from assembly language code must be listed in "sys/i386/include/asnames.h". It so happens that failure to do these things will have no impact on the a.out kernel. But it will break the build of the ELF kernel. The ELF bootloader works, but it is not ready to commit quite yet.
1997-04-22 06:55:47 +00:00
${NORMAL_C}
Use the gensetdefs that is now in /usr/bin, it's a superset of the one in the kernel source, and that one is already used for modules. I don't _think_ this will hurt releases, aout-to-elf, etc, but it is possible. In all the cases I've looked at, config(8) has been generated straight after a make world, so if /usr/sbin/config exists and is the right version for the kernel, then we can pretty much count on /usr/bin/gensetdefs being there too.
1999-01-17 21:17:27 +00:00
setdef1.o: setdef1.c setdefs.h
Make the necessary changes so that an ELF kernel can be built. I have successfully built, booted, and run a number of different ELF kernel configurations, including GENERIC. LINT also builds and links cleanly, though I have not tried to boot it. The impact on developers is virtually nil, except for two things. All linker sets that might possibly be present in the kernel must be listed in "sys/i386/i386/setdefs.h". And all C symbols that are also referenced from assembly language code must be listed in "sys/i386/include/asnames.h". It so happens that failure to do these things will have no impact on the a.out kernel. But it will break the build of the ELF kernel. The ELF bootloader works, but it is not ready to commit quite yet.
1997-04-22 06:55:47 +00:00
${NORMAL_C}
Use the gensetdefs that is now in /usr/bin, it's a superset of the one in the kernel source, and that one is already used for modules. I don't _think_ this will hurt releases, aout-to-elf, etc, but it is possible. In all the cases I've looked at, config(8) has been generated straight after a make world, so if /usr/sbin/config exists and is the right version for the kernel, then we can pretty much count on /usr/bin/gensetdefs being there too.
1999-01-17 21:17:27 +00:00
setdef0.c setdef1.c setdefs.h: ${OBJS}
Hide the gensetdefs run the same way we hide the final ld command.
1999-01-19 17:08:28 +00:00
@gensetdefs ${OBJS}
This commit affects ELF kernels only. Remove "setdefs.h" and arrange to generate it automatically at ELF kernel build time. "gensetdefs.c" is a utility which scans a set of ELF object files and outputs a line ``DEFINE_SET(name, length);'' for each linker set that it finds. When generating an ELF kernel, this is run just before the final link to generate "setdefs.h". Remove the init_sets() function from "setdef0.c", and its call from "machdep.c". Since "gensetdefs.c" calculates the length of each set, it is no longer necessary in an ELF kernel to count the set elements at kernel initialization time. Also remove "set_of_sets" which was used for this purpose. Link "setdef0" and "setdef1" into the kernel only if building for ELF. Since init_sets() is no longer used, there is no need to link them into an a.out kernel.
1997-05-21 23:21:30 +00:00
Remove exception.o, swtch.o, support.o, conf.o and machdep.o from SYSTEM_OBS. They are now normal objects. Remove stale dependencies for the above now-normal objects and for locore.o and generate dependencies using mkdep. Config doesn't generate lists of assembler source files so the lists to be mkdep'ed have to be given explictly. Only the standard *.s files are given, so the dependencies for gnu/fpemul/*.s are incomplete. *.S files would be handled right if config put them in CFILES. Don't define NPX. It was replaced by NNPX > 0 years ago. Define LOAD_ADDRESS in COPTS so that compiling machdep.c isn't a special case. Moving around the dependencies exposed a bug in make. It doesn't know that assym.s and ./assym.s are the same. Add a rule tell it.
1994-08-30 17:17:49 +00:00
# this rule stops ./assym.s in .depend from causing problems
./assym.s: assym.s
Removed un-necessary dependencies for assym.s
1993-10-23 08:28:17 +00:00
assym.s: genassym
Never strip the kernel. Recommend -Wimplicit in CWARNFLAGS next. There are still a few hundred potential arg mismatches because no function declaration is in scope. Don't duplicate option `-I.'. Remove null editing of the assembler source for all profiled objects. The required magic has been done since prehistoric times by an asm("mcount") declaration. Simplify the clean rule. Don't try to be clever about timestamps involving genassym. genassym's timestamp usually got ahead of assym.s's timestamp, so `make' almost always had to run genassym and compare *assym.s to decide that nothing needed to be done. The cost is reassembling a few files whenever genassym is rebuilt. Assembling is almost as fast as comparing. Always go through genassym.o to build genassym. This would have avoided numerous bugs involving mkdep -p. Now it just stops genassym from depending on the name of the temporary object file. Use ${CFLAGS} for building genassym. Mainly ${CWARNFLAGS} were missing.
1994-10-25 19:34:57 +00:00
./genassym >assym.s
Try to be a little smarter about touching the date on assym.s, so that if something changes which doesn't affect it, locore doesn't have to get rebuilt. This is at the cost of a genassym and a cmp in every compile, until someone can figure out how to make `make' smarter itself.
1993-11-07 16:46:33 +00:00
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
genassym.o: ${I386}/i386/genassym.c
Fixed missing dependencies of genassym.o, gensetdefs.o and vers.o on opt_global.h. This actually matters for genassym.o (it depends on at least SMP). Don't undefine KERNEL for compiling genassym.c. genassym.o really depends on KERNEL, and the prototype mismatches that required undefining KERNEL in rev.1.49 no longer exist. Compile gensetdefs* with the same flags as genassym*. External `gen' programs such linux_genassym should also use these flags (${GEN_CFLAGS}). Fixed missing dependency generation for gensetdefs.o. C sources for external `gen' programs should be added to GEN_CFILES to get their dependencies generated. Cleaned up flags definitions and use. All of ${CFLAGS} is now passed to mkdep and the assembler (both are actually variants of ${CC} and will ignore the irrelevant flags).
1998-09-21 06:39:11 +00:00
${CC} -c ${GEN_CFLAGS} ${I386}/i386/genassym.c
Never strip the kernel. Recommend -Wimplicit in CWARNFLAGS next. There are still a few hundred potential arg mismatches because no function declaration is in scope. Don't duplicate option `-I.'. Remove null editing of the assembler source for all profiled objects. The required magic has been done since prehistoric times by an asm("mcount") declaration. Simplify the clean rule. Don't try to be clever about timestamps involving genassym. genassym's timestamp usually got ahead of assym.s's timestamp, so `make' almost always had to run genassym and compare *assym.s to decide that nothing needed to be done. The cost is reassembling a few files whenever genassym is rebuilt. Assembling is almost as fast as comparing. Always go through genassym.o to build genassym. This would have avoided numerous bugs involving mkdep -p. Now it just stops genassym from depending on the name of the temporary object file. Use ${CFLAGS} for building genassym. Mainly ${CWARNFLAGS} were missing.
1994-10-25 19:34:57 +00:00
genassym: genassym.o
Decouple genassym flags from CFLAGS. Submitted by: jhay
1998-09-14 11:32:17 +00:00
${CC} ${GEN_CFLAGS} genassym.o -o ${.TARGET}
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Use the gensetdefs that is now in /usr/bin, it's a superset of the one in the kernel source, and that one is already used for modules. I don't _think_ this will hurt releases, aout-to-elf, etc, but it is possible. In all the cases I've looked at, config(8) has been generated straight after a make world, so if /usr/sbin/config exists and is the right version for the kernel, then we can pretty much count on /usr/bin/gensetdefs being there too.
1999-01-17 21:17:27 +00:00
${SYSTEM_OBJS} genassym.o vers.o: opt_global.h
Create a special option file "opt_global.h" which is included by all source files via a 'cc -include opt_global.h ...' type arrangement. This means we can untangle certain header files. options.i386 has a placeholder until it has a real member so we can avoid having to teach config about it just yet. Reviewed by: bde
1997-04-27 20:01:47 +00:00
Handle the aic7770 driver's dependancies correctly. YOU MUST REBUILD CONFIG.
1994-12-31 19:27:29 +00:00
depend: assym.s param.c vnode_if.h ${BEFORE_DEPEND}
When doing a 'make depend', dont blow away the existing .depend until the generation of a new one is complete.
1996-03-30 14:57:02 +00:00
rm -f .newdep
Fixed missing dependencies of genassym.o, gensetdefs.o and vers.o on opt_global.h. This actually matters for genassym.o (it depends on at least SMP). Don't undefine KERNEL for compiling genassym.c. genassym.o really depends on KERNEL, and the prototype mismatches that required undefining KERNEL in rev.1.49 no longer exist. Compile gensetdefs* with the same flags as genassym*. External `gen' programs such linux_genassym should also use these flags (${GEN_CFLAGS}). Fixed missing dependency generation for gensetdefs.o. C sources for external `gen' programs should be added to GEN_CFILES to get their dependencies generated. Cleaned up flags definitions and use. All of ${CFLAGS} is now passed to mkdep and the assembler (both are actually variants of ${CC} and will ignore the irrelevant flags).
1998-09-21 06:39:11 +00:00
mkdep -a -f .newdep ${CFLAGS} ${CFILES} ${SYSTEM_CFILES}
mkdep -a -f .newdep ${GEN_CFLAGS} ${GEN_CFILES}
env MKDEP_CPP="${CC} -E" \
mkdep -a -f .newdep ${ASM_CFLAGS} ${SFILES} ${SYSTEM_SFILES}
When doing a 'make depend', dont blow away the existing .depend until the generation of a new one is complete.
1996-03-30 14:57:02 +00:00
rm -f .depend
``mv'' -> ``mv -f'' ``rm'' -> ``rm -f'' so mv/rm may not ask for confirmation if you are not root
1996-05-07 23:19:49 +00:00
mv -f .newdep .depend
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Add support for cleandepend, ${DESTDIR}, and there not being a file named ${DESTDIR}/kernel.
1998-01-31 04:23:15 +00:00
cleandepend:
rm -f .depend
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
links:
Move a couple of bits of (now) static content from config(8) (ie: %LOAD) directly into the Makefile. Remove references to swapkernel.c, it's not generated by config(8) now. (The previous config commits had generated it, but they had an unused 'char *' in them).
1999-05-09 18:57:02 +00:00
egrep '#if' ${CFILES} | sed -f $S/conf/defines | \
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
sed -e 's/:.*//' -e 's/\.c/.o/' | sort -u > dontlink
Move a couple of bits of (now) static content from config(8) (ie: %LOAD) directly into the Makefile. Remove references to swapkernel.c, it's not generated by config(8) now. (The previous config commits had generated it, but they had an unused 'char *' in them).
1999-05-09 18:57:02 +00:00
echo ${CFILES} | tr -s ' ' '\12' | sed 's/\.c/.o/' | \
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
sort -u | comm -23 - dontlink | \
sed 's,../.*/\(.*.o\),rm -f \1;ln -s ../GENERIC/\1 \1,' > makelinks
sh makelinks && rm -f dontlink
tags:
Working kernel tags! Submitted by: Craig Leres <leres@ee.lbl.gov> PR: 2806
1999-07-02 04:00:01 +00:00
@[ -f .depend ] || { echo "you must make depend first"; exit 1; }
sh $S/conf/systags.sh
rm -f tags1
sed -e 's, ../, ,' tags > tags1
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Clean up the kernel/kernel.debug/install.debug etc stuff. Sense ${DEBUG} to decide on building kernel.debug or not. Use a common install and install.debug target to minimize duplication. (I deleted the ELF transition kernel warning, it was getting dated..)
1999-04-13 18:25:08 +00:00
install install.debug:
@if [ ! -f ${KERNEL}${.TARGET:S/install//} ] ; then \
Typo
1999-11-16 22:02:16 +00:00
echo "You must build a kernel first." ; \
Replace the dependency of `install' from `kernel' by a test for the existance fo `kernel'. Submitted by: peter
1996-05-02 21:40:50 +00:00
exit 1 ; \
fi
1. Modify config to issue different code for debugging. 2. Config complains if you use -g: Debugging is enabled by default, there is no ned to specify the -g option 3. Config warns you if you don't use -s: Building kernel with full debugging symbols. Do "config -s BSD" for historic partial symbolic support. To install the debugging kernel, do make install.debug (BSD was the name of the config file I used; I print out the same name). 4. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 and config to work if a kernel name other than 'kernel' is specified. This is not absolutely necessary, but useful, and it was relatively easy. I now have a kernel called /crapshit :-) 5. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 "clean" target to remove both the debug and normal kernel. 6. Modify all to install the stripped kernel by default and the debug kernel if you enter "make install.debug". 7. Update version number of Makefiles and config.
1999-04-07 09:28:03 +00:00
.if exists(${DESTDIR}/${KERNEL})
-chflags noschg ${DESTDIR}/${KERNEL}
mv ${DESTDIR}/${KERNEL} ${DESTDIR}/${KERNEL}.old
Add support for cleandepend, ${DESTDIR}, and there not being a file named ${DESTDIR}/kernel.
1998-01-31 04:23:15 +00:00
.endif
Augment $PATH to ensure searching of /sbin and /usr/sbin for sysctl instead of using an absolute path to sysctl. Problem reported by: ache
1997-12-18 21:03:04 +00:00
PATH=$${PATH}:/sbin:/usr/sbin; \
1. Modify config to issue different code for debugging. 2. Config complains if you use -g: Debugging is enabled by default, there is no ned to specify the -g option 3. Config warns you if you don't use -s: Building kernel with full debugging symbols. Do "config -s BSD" for historic partial symbolic support. To install the debugging kernel, do make install.debug (BSD was the name of the config file I used; I print out the same name). 4. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 and config to work if a kernel name other than 'kernel' is specified. This is not absolutely necessary, but useful, and it was relatively easy. I now have a kernel called /crapshit :-) 5. Modify Makefile.i386, Makefile.alpha, Makefile.pc98 "clean" target to remove both the debug and normal kernel. 6. Modify all to install the stripped kernel by default and the debug kernel if you enter "make install.debug". 7. Update version number of Makefiles and config.
1999-04-07 09:28:03 +00:00
if [ `sysctl -n kern.bootfile` = ${DESTDIR}/${KERNEL} ] ; then \
sysctl -w kern.bootfile=${DESTDIR}/${KERNEL}.old ; \
Move KERNEL?=kernel to top of the file where it's more obvious and add a pointer to 'makeoptions' and /etc/make.conf. Catch a few stray "kernel" hardcoded references. Move the kernel.debug and related kernel build rules together.
1999-05-09 22:42:17 +00:00
if [ -f /var/db/kvm_${KERNEL}.db ] ; then \
mv -f /var/db/kvm_${KERNEL}.db /var/db/kvm_${KERNEL}.old.db ; \
Test if kvm_kernel.db exists before moving it. Otherwise installing a kernel, booting single user and reinstalling a kernel fails.
1997-11-03 21:48:31 +00:00
fi \
Only change the sysctl variable if it presently points to /kernel.
1996-01-07 22:31:27 +00:00
fi
Clean up the kernel/kernel.debug/install.debug etc stuff. Sense ${DEBUG} to decide on building kernel.debug or not. Use a common install and install.debug target to minimize duplication. (I deleted the ELF transition kernel warning, it was getting dated..)
1999-04-13 18:25:08 +00:00
install -c -m 555 -o root -g wheel -fschg \
${KERNEL}${.TARGET:S/install//} ${DESTDIR}/${KERNEL}
Don't include options MULTICAST or MROUTING in generic kernels. MULTICAST has no effect now, and MROUTING should never be defined by default. (Eventually the code should be dynamically loadable.) Also, allow for Pentium CPUs in GENERICBT kernels. Add an `install' rule to Makefile.i386, which looks like this: mv /kernel /kernel.old install -c -m 555 -o root -g root -fschg kernel / I'd like comments on whether or not you think it's a good idea to have the kernel be immutable by default; I'm happy either way. CVS:
1994-09-09 22:03:51 +00:00
I'm sick of the automatic rename of /kernel to /kernel.old while doing development that leads to lots of crashes during boot. I have made a 'reinstall' target (like in ports, and reinstall.debug) This is most useful if you want to keep /kernel.old as a known bootable kernel. If you test a new kernel and have to reboot for a fix, a 'make reinstall' will install the new kernel over the top of the old non-viable one, leaving the old one untouched. This is mainly meant for development, not general users.
1999-05-08 20:04:39 +00:00
reinstall reinstall.debug:
install -c -m 555 -o root -g wheel -fschg \
${KERNEL}${.TARGET:S/reinstall//} ${DESTDIR}/${KERNEL}
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
config.o:
${NORMAL_C}
ioconf.o:
${NORMAL_C}
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
This Makefile is for FreeBSD, not 4.3 BSD-Reno. Moved $Id$. Added STRIP=, DBSYM=, and LOAD_ADDRESS?= Now use LOAD_ADDRESS for linking kernel and for dbsym, added strip -x to cut kernel size. Added machde.o: dependency, this will be needed in the future, and for now it does not hurt anyone. Cleaned out conf.o: dependency, mkdep does the right things. Same for param.c: This is really a Merge in of NetBSD's Makefile.i386, here is the relevant rlog info: ---------------------------- revision 1.27 date: 1993/08/27 23:58:20; author: brezak; state: Exp; lines: +2 -2 Need LOAD_ADDRESS for depend pass. ---------------------------- revision 1.25 date: 1993/07/19 16:52:16; author: mycroft; state: Exp; lines: +3 -3 Add ${DEBUG} to CFLAGS and -f to dbsym. ---------------------------- revision 1.22 date: 1993/07/18 10:08:22; author: mycroft; state: Exp; lines: +5 -6 Change to work with new config stuff for specifying load address. ---------------------------- revision 1.20 date: 1993/07/18 09:47:40; author: mycroft; state: Exp; lines: +6 -5 Use new -T option to dbsym. ---------------------------- revision 1.17 date: 1993/07/11 08:42:22; author: cgd; state: Exp; lines: +2 -2 don't ignore errors from dbsym... it might say that, e.g. there's not enough symbol space! ---------------------------- revision 1.14 date: 1993/06/06 23:29:03; author: cgd; state: Exp; lines: +2 -2 make conf.o actually depend on conf.c... ---------------------------- revision 1.8 date: 1993/04/29 03:27:39; author: cgd; state: Exp; lines: +5 -10 use ed instead of ex. the script to use is identical, and we might want to switch back to using ex when our ex supports -. ---------------------------- revision 1.5 date: 1993/03/24 18:48:57; author: cgd; state: Exp; lines: +1 -1 now use absolute path for dbsym ----------------------------
1993-09-26 19:44:33 +00:00
param.c: $S/conf/param.c
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
-rm -f param.c
cp $S/conf/param.c .
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
param.o:
${NORMAL_C}
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
vers.c: $S/conf/newvers.sh $S/sys/param.h ${SYSTEM_DEP}
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
sh $S/conf/newvers.sh ${KERN_IDENT} ${IDENT}
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
# XXX strictly, everything depends on Makefile because changes to ${PROF}
# only appear there, but we don't handle that.
vers.o:
${NORMAL_C}
Fix race condition for creation of vnode_if.c and vnode_if.h. Exact problem pointed out by: Cyrus Rahman <cr@jcmax.com>
1998-06-24 13:55:32 +00:00
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
.ORDER: vnode_if.c vnode_if.h
vnode_if.c vnode_if.h: $S/kern/vnode_if.sh $S/kern/vnode_if.src
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
sh $S/kern/vnode_if.sh $S/kern/vnode_if.src
Removed historical dependencies on `Makefile'. They had rotted to being mostly for objects that have the fewest dependencies on `Makefile' (since they were mostly for utilities and objects generated from *.s and these don't depend on profiling flags). Give an explicit rule for building vnode_if.o. This fixes building it without ${PROF}. Use .ORDER instead of a stamp file to avoid building vnode_if.[ch] concurrently. Removed explicit dependencies that will be generated by `make' (.c.o) or will be generated by mkdep. Added missing dependencies of special objects on opt_global.h. Use ${NORMAL_C} instead of special rules for special objects where possible. FIxed dependencies of vers.o.
1998-07-12 10:47:32 +00:00
vnode_if.o:
${NORMAL_C}
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
Hack: comment out the pull-in of src/sys/../share/mk/bsd.kern.mk as that breaks if you try and compile a kernel before building world, as is presently required to get past the signal changes. I don't particularly like doing this, but at least it will mean that a 'make world' will activate the gcc 2.95.2-specific option in bsd.kern.mk.
1999-11-18 09:54:11 +00:00
# Commented out for now pending a better solution.
# How do we pick up compiler version specific flags??
#.if exists($S/../share/mk)
#.include "$S/../share/mk/bsd.kern.mk"
#.else
Centralized the definition of CWARNFLAGS into bsd.kern.mk.
1996-10-08 22:09:03 +00:00
.include <bsd.kern.mk>
Hack: comment out the pull-in of src/sys/../share/mk/bsd.kern.mk as that breaks if you try and compile a kernel before building world, as is presently required to get past the signal changes. I don't particularly like doing this, but at least it will mean that a 'make world' will activate the gcc 2.95.2-specific option in bsd.kern.mk.
1999-11-18 09:54:11 +00:00
#.endif
Centralized the definition of CWARNFLAGS into bsd.kern.mk.
1996-10-08 22:09:03 +00:00
Initial import, 0.1 + pk 0.2.4-B1
1993-06-12 14:58:17 +00:00
%RULES
# DO NOT DELETE THIS LINE -- make depend uses it
Reference in New Issue Copy Permalink