Commit Graph

193 Commits

Author SHA1 Message Date
ru
8735fdbd4c Enable GCC stack protection (aka Propolice) for userland:
- It is opt-out for now so as to give it maximum testing, but it may be
  turned opt-in for stable branches depending on the consensus.  You
  can turn it off with WITHOUT_SSP.
- WITHOUT_SSP was previously used to disable the build of GNU libssp.
  It is harmless to steal the knob as SSP symbols have been provided
  by libc for a long time, GNU libssp should not have been much used.
- SSP is disabled in a few corners such as system bootstrap programs
  (sys/boot), process bootstrap code (rtld, csu) and SSP symbols themselves.
- It should be safe to use -fstack-protector-all to build world, however
  libc will be automatically downgraded to -fstack-protector because it
  breaks rtld otherwise.
- This option is unavailable on ia64.

Enable GCC stack protection (aka Propolice) for kernel:
- It is opt-out for now so as to give it maximum testing.
- Do not compile your kernel with -fstack-protector-all, it won't work.

Submitted by:	Jeremie Le Hen <jeremie@le-hen.org>
2008-06-25 21:33:28 +00:00
jb
f0232b81e2 Re-enable -Werror again.
This time, change the default CFLAGS to match the simple defaults that
the tinderboxes use. That is, don't use -fno-strict-aliasing by default.

My last attempt to re-anable -Werror gave me a lesson in what strict
aliasing is all about. There was code in libthr that wasn't 64-bit clean.
The default use of -fno-strict-aliasing hid that.

By using -fno-strict-aliasing by default we were choosing to ignore
problems in code which had the potential to shoot ourselves in the
foot. Sometimes it would be the 64-bit foot. I have both feet. The left
ones are 32 bits and the right ones are 64 bits. Don't ask about my
endian orientation. :-)

The -fno-strict-aliasing compiler arg can still be used if NO_STRING_ALIASING
is define in make.

We are early in the FreeBSD 8 development, so we have the opportunity to
wait and see if this works for us. I am sure that people will complain.
We can easily revert this. All I ask is that we take sides: clean code or
not. YMMV.

Note that by using -fno-strict-aliasing the build won't actually break.
Only where WARNS is set (and -Werror is used) will a compiler warning break
the build. The use of WARNS levels implies (to me at least) that the
developer has taken some care to make the code pass basic checks. This
commit makes those checks just a little bit more strict.
2007-11-22 23:21:12 +00:00
jb
f2dddc3de3 Unfortunately the tinderbox setup uses custom CFLAGS which are a big
obstacle to enabling -Werror. I'll continue to work on cleaning up the
code so that we can keep this enabled.

If the tinderboxes would just use the default CFLAGS set in this file,
all would be fine and we'd be able to make use of -Werror.
2007-11-19 22:39:39 +00:00
jb
f8ce423552 Re-enable -Werror ins WARNS as it was roughly 6 months ago before
being disabled while gcc 4.2 was bedded in.

Tested with 'make release' (amd64 arm i386 ia64 pc98 powerpc sparc64 sun4v)
2007-11-19 09:09:02 +00:00
obrien
ae19feef0a Temporarily add 'WITH_GCC3' that removes -Wno-pointer-sign from the
compiler invocation.  This is just to help get over the hump of people
tracking down bugs that may cross the GCC 4.2 upgrade.
It is envisioned that this option goes away after a suitable amount
of time.
2007-05-24 21:53:42 +00:00
kan
925e7e54c5 Universally disable -Werror until src/ is in better shape for GCC 4.2.
There are new warnings that kill the build otherwise.

Disable pointer destination sign mismatch warning alltogether. Our tree
is in no shape to have that enabled yet.
2007-05-19 04:41:05 +00:00
jb
d871054da8 Allow a makefile to set IGNORE_PRAGMA so that OpenSolaris code can
be built with other gcc warnings enabled.

Every Solaris source file has a #pragma ident in it. We can just
ignore those definitions.
2006-11-04 04:39:05 +00:00
des
be4056d8d7 Reintroduce CSTD, which allows a Makefile to specify the precise dialect
of C in which the program or library is written.

Note that this is *not* intended to be used across the whole tree.  It
is intended to be used for individual libraries or programs which use
specific language features which the compiler must know about in order
to produce correct warnings at high WARNS levels.

MFC after:	1 month
2006-08-11 17:28:59 +00:00
obrien
a8c5c0fe7e Move -Wunused-parameter from WARNS level 3 level 4.
Also break long lines -- note that the '\' must be up against the last
character of a line to keep command-line spacing proper.

Requested by:	rwatson
2005-01-16 21:18:16 +00:00
obrien
5568092622 Use >= so we can use the actual WARNS levels. 2005-01-16 21:08:31 +00:00
trhodes
1492567af1 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
des
bc27cba087 Remove -Wbad-function-cast. Its main purpose is to catch bugs that we
already catch with -Wstrict-prototypes, and it causes spurious warnings
for some perfectly legitimate constructs.
2004-02-16 20:07:06 +00:00
ru
032a5ac8ac Put the warning flags to where they belong (into CWARNFLAGS).
This allows us to easily override them when necessary, e.g.,
to selectively disable warnings in libc/ contributed sources.
2004-01-11 10:29:55 +00:00
imp
93442457e1 Put on the core hat and back out all of the CSTD= changes. Core will
deal with working with the parties to define a coherent definition for
CSTD that doesn't break things.

Core hat seconded by: markm
2003-06-14 17:41:59 +00:00
des
5d547700c7 Revert to a known-good state. Anyone desiring to experiment with stricter
global settings is free to do so in his or her own source tree.
2003-06-14 11:57:44 +00:00
peter
bc201df6b8 We cannot use c99 on amd64 either due to lack of alloca(). libc:strptime()
uses alloca() and alloca is impossible to implement as a callable function
on amd64.  It has to be a compiler builtin.  Note that the bigger problem
is that libc is not c99 clean internally.
2003-06-13 21:54:21 +00:00
obrien
9b92f918c2 Be C std strict on i386 and amd64 as we can. Be loose on Alpha and ia64. 2003-06-07 08:05:35 +00:00
obrien
93e4e9df9b Compile our code as C99 w/GNU extensions by default.
We can't use straight "c99" due to the lack of alloca.S for non-i386 platforms.
2003-06-06 16:55:05 +00:00
obrien
5c52044b3d I got a bazzar bug report 2003-06-02 08:10:57 +00:00
obrien
4f1dd109b8 Turn back on c99, the tree should be ready for it now. 2003-06-02 06:26:14 +00:00
obrien
c2c5b39ab7 To quote Tony Maher <tonymaher@optushome.com.au>, "maybe 3rd time lucky ;-)"
*sigh* Just can't get a brake when trying to react too quickly and fix the build.
2003-06-02 01:01:42 +00:00
obrien
e15b452288 Temperarly turn off building the tree with c99.
I swore this made it thru a 'make world', but I don't know what happened.
2003-06-02 00:11:07 +00:00
obrien
45f07c7359 Use a bigger hammer -- keep all -std= out of CXXFLAGS.
Also allow for "CSTD=" in a Makefile.
2003-06-01 22:13:45 +00:00
obrien
00cddc9c66 Do not set a C standard for the C++ compiler. 2003-06-01 21:33:05 +00:00
obrien
07f2e845ae Compile our code as C99 by default. 2003-06-01 18:31:25 +00:00
obrien
8c39502975 CSTD is virtually worthless for WARNS=2-5. Return -pedantic to WARNS=6+. 2003-05-31 16:37:22 +00:00
obrien
8b0354b4be Make CSTD style match the rest of file. 2003-05-31 16:30:39 +00:00
des
495eda8ee4 Whitespace cleanup (1.15 had spaces instead of tabs) 2003-05-05 21:14:36 +00:00
des
2a6383574f Don't use -pedantic unless we also set -std of some kind. 2003-05-05 21:13:32 +00:00
des
d147648191 Add a mechanism to allow Makefiles to specify the particular C dialect
in which the source code is written.  This is controlled by the CSTD
variable, which can have one of the following values:

  - "k&r"		=> -traditional
  - "c89" or "c90"	=> -std=iso9899:1990
  - "c94" or "c95"	=> -std=iso9899:199409
  - "c99"		=> -std=iso9899:1999

The corresponding option is added to CFLAGS regardless of WARNS level.
This also removes -ansi from WARNS level 6, but adds -Wno-long-long to
work around a weird gcc bug (-ansi, which is supposedly equivalent to
-std=iso9899:1990, seems to turn long long warnings off instead of on)

If CSTD is undefined, CFLAGS are unchanged except for the -ansi /
-Wno-long-long change mentioned above for WARNS level 6.
2003-05-01 14:36:27 +00:00
imp
b0693a4cbc Migrate to a new way of dealing with building from old revisions of
FreeBSD.  This method attempts to centralize all the necessary hacks
or work arounds in one of two places in the tree (src/Makefile.inc1
and src/tools/build).  We build a small compatibility library
(libbuild.a) as well as selectively installing necessary include
files.  We then include this directory when building host binaries.

This removes all the past release compatibilty hacks from various
places in the tree.  We still build on tip of stable and current.  I
will work with those that want to support more, although I anticipate
it will just work.

Many thanks to ru@, obrien@ and jhb@ for providing valuable input at
various stage of implementation, as well as for working together to
positively effect a change for the better.
2003-04-05 20:30:30 +00:00
ru
db5ec15123 -Wall implies -Wuninitialized if -O is also in effect.
-Wuninitialized does not work without -O.

This fixes the ${WARNS} > 4 compilations with -O0.

Spotted by:	marcel
2003-04-04 10:47:06 +00:00
ru
874869c32c Enable cpp(1) warnings in system headers. GCC is oriented on
glibc which is externally maintained, so GCC ships with these
warnings turned off by default.  This is also consistent with
the src/contrib/gcc/c-lex.c,v 1.2 change.
2003-03-31 13:10:51 +00:00
ru
7920e22cea Take __FreeBSD_version into account when BOOTSTRAPPING. 2002-11-13 13:49:29 +00:00
obrien
6793e7aaf6 Be more strict with WARNS now -- the GCC 3.1 turmoil is behind us. 2002-07-07 18:47:52 +00:00
obrien
48306d2934 Tweak the WARNS levels a tad. 2002-05-19 18:24:00 +00:00
obrien
4267f26cd1 Gcc 3.1 has different -Wx flags. 2002-05-10 01:58:16 +00:00
obrien
824b21ae00 Add WARNS levels 5 and 6:
5 gives us -Wuninitialized rather than -Wno-uninitialized
6 gives us full BDECFLAGS
2002-04-10 02:45:22 +00:00
ru
b42f3eadf7 FreeBSD prior to 4.5-RELEASE and older versions of 5.0-CURRENT do not
have the __FBSDID() macro in <sys/cdefs.h>.  Fix this once and for all
for tools that need to be bootstrapped.

PR:		bin/36747
MFC after:	3 days
Prodded by:	obrien
2002-04-07 17:03:06 +00:00
dwmalone
ff2a69a53b Put -Wno-uninitialized at the end of the warns flags as it may be enabeled
by several flags (in this case -W and -Wall).

Reviewed by:	bde
2002-01-05 20:17:50 +00:00
obrien
43ede8b147 Add some granularity to the WARNS levels.
1:  add -Werror
2:  -Wall [only], as this is the most used warnings setting by developers
3:  our old `1'
4:  our old `2'
2001-12-15 06:02:15 +00:00
kris
4029cc2464 Only add -Werror if in fact we have set WARNS or WFORMAT to >0 2001-05-19 23:32:19 +00:00
kris
3252dc5107 Bring in the WARNS directive from NetBSD, localized in a new bsd.sys.mk
to avoid polluting sys.mk.  This directive controls the addition of
compiler warning flags to CFLAGS in a relatively compiler-neutral manner.

The idea is that WARNS can be set in Makefile.inc or in individual
Makefiles as they become clean, to prevent the introduction of new
warnings in the code.  -Werror is added by default
2001-05-19 23:18:21 +00:00