Commit graph

350 commits

Author SHA1 Message Date
Tim J. Robbins fadfe1d425 Test for bug in revision 1.54 of vfprintf.c. 2003-04-06 08:02:20 +00:00
Warner Losh f7e19a98fd We can't use ${WORLDTMP}/usr/{include,lib} for the compat layer. This
is because we populate these directories later, and a subsequent
-DNOCLEAN build may fail.  So, we put them in
${WORLDTMP}/build/usr/{include,lib} instead and adjust Makefile.boot.

Again, this works on -stable and -current, but might break older
versions.

Submitted by: ru@
2003-04-06 03:50:28 +00:00
Warner Losh 5222688930 Don't need the gccism include_next since we won't create this file
when a real sys/endian.h exists.

Submitted by: ru@
2003-04-06 03:30:44 +00:00
David Schultz 600276e186 Regression tests for printf()'s floating point formats.
Tests for correct printing of subnormals are lacking.
2003-04-05 22:13:01 +00:00
Warner Losh 30aaff1192 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
Warner Losh 2949da0a2e 2.x->3.0 upgrade tools are no longer needed. 5.x is unlikely to fit
on a 2.1/2.2 root disk anyway.

OK'd by: peter
2003-04-03 05:44:08 +00:00
David Schultz 0c66f5adc3 Write negative zero as '-0.0'. Otherwise, it is parsed as an integer
before being cast to a floating point type, and the sign is lost.
2003-03-27 05:32:28 +00:00
Sam Leffler eabb18a6fc add missing %u to show mcr1full count 2003-03-24 17:36:50 +00:00
David Xu 430bd13fe9 Follow new upcall quantum code in kernel. 2003-03-21 08:21:39 +00:00
Sam Leffler 46cdb828e6 o add missing copyright notices
o new hifnstats program to dump hifn driver statistics
o new ipsecstats program to dump fast ipsec statistics
2003-03-19 01:53:55 +00:00
Dag-Erling Smørgrav ee5803b89b Allow the configuration to specify environment variables (passed on
the command line to tinderbox.pl)
Build Kerberos V in all setups (this was previously taken care of by
tinderbox.pl).
Tweak the 9ball configuration to make powerpc builds work (with a
little help from a toolchain patch provided by grehan)
2003-03-14 00:55:41 +00:00
Dag-Erling Smørgrav 9c48602d6c Support multiple verbosity levels.
Allow the user to specify environment variables on the command line.
Set the correct variables for cross-building.
Don't arbitrarily define MAKE_KERBEROS5 and BOOT_UFS.
2003-03-14 00:52:54 +00:00
Sam Leffler a0e32c7bc6 use relative pathname to driver-private file (instead of absolute) 2003-03-11 18:49:50 +00:00
Sam Leffler 3bf663f5f1 add simple program to dump ubsec driver statistics 2003-03-11 18:42:20 +00:00
Mark Murray b7d3fc8d5b KerberosIV deorbit sequence: Remove references to KerbIV libs. 2003-03-08 12:14:26 +00:00
Dag-Erling Smørgrav c13dcea6ad Prepare for removal of KerberosIV from the tree. 2003-03-06 17:13:18 +00:00
Poul-Henning Kamp 49efad87df Fix a bug in the white-space remover. 2003-03-04 19:20:52 +00:00
David Xu ec2af999d3 Add test program for round robin in userland. 2003-02-26 01:05:10 +00:00
Dag-Erling Smørgrav 473ef84576 Hacks to get i386 and pc98 building:
- TARGET=${TARGET_MACHINE}
 - BOOT2_UFS=UFS1_ONLY
2003-02-24 10:48:01 +00:00
Sam Leffler c17b39c12e o add -b option to enable batching
o add #ifdef __FreeBSD__'s around the profiling so the code compiles on openbsd
o update a comment about expected performance for Broadcom hardware
2003-02-24 06:10:50 +00:00
Dag-Erling Smørgrav bbcbd93649 If more than one argument were given on the command line, interpret the
additional arguments as a series of architectures to build.  This is
useful for running subsets of a particular configuration.
2003-02-23 21:24:08 +00:00
Dag-Erling Smørgrav 3ceb962db6 Add LINT to the target list for the test setup. 2003-02-23 12:41:32 +00:00
Dag-Erling Smørgrav db08524b17 Add LINT to the taget list for CURRENT. 2003-02-23 12:40:21 +00:00
Dag-Erling Smørgrav 7941c77b0a Don't try to build LINT if there is no NOTES file. 2003-02-23 12:39:25 +00:00
Dag-Erling Smørgrav 249834de3d Don't omit --update on 9ball. 2003-02-21 07:33:41 +00:00
David Xu 5f22df9792 Adjust code for new kse_release interface. 2003-02-20 08:24:22 +00:00
Dag-Erling Smørgrav 51260bb27f Properly terminate some warning messages. 2003-02-18 12:50:37 +00:00
Dag-Erling Smørgrav b61b86b86f Explicitly set TZ to GMT (including in the build environment) 2003-02-18 10:14:13 +00:00
Dag-Erling Smørgrav 9dc3de8830 - build powerpc world on 9ball
- log directly into two files: a full log and a summary
- set timezone to GMT before starting
2003-02-17 19:50:58 +00:00
Dag-Erling Smørgrav e52af12a16 Log to stdout if no file was specified. 2003-02-15 19:32:52 +00:00
Dag-Erling Smørgrav 374053b528 Avoid an undef warning. 2003-02-15 14:30:51 +00:00
Dag-Erling Smørgrav fbacf8a2cc Decrease verbosity, fix some configuration problems. 2003-02-14 09:18:44 +00:00
Dag-Erling Smørgrav 55c4f5c07a Make the tinderbox log lines less obtrusive. 2003-02-13 21:21:48 +00:00
Dag-Erling Smørgrav 0e01af7032 Pass KERNCONF on the command line rather than in the environment,
since the command line is included in the log.
2003-02-13 21:19:20 +00:00
Dag-Erling Smørgrav 9929cb456d Complete overhaul of the configuration system, allowing the user to
select one of multiple configurations at runtime.
2003-02-13 21:18:19 +00:00
Dag-Erling Smørgrav 091aeb5c3c GC the old tinderbox script. 2003-02-13 14:27:11 +00:00
Dag-Erling Smørgrav 902f48ca91 Place the logs in the tinderbox directory rather than cwd. 2003-02-13 11:57:13 +00:00
Dag-Erling Smørgrav 64e31b67e1 Move away the old log file before starting the tinderbox so we don't
accidentally post the old log if the tinderbox script failed to start
altogether.
2003-02-13 10:52:46 +00:00
Dag-Erling Smørgrav 72d8c7ffcd - run the tinderbox script in verbose mode
- use /\bStop\b/ instead of /^\*\*\* Error/ to identify errors
2003-02-13 09:59:59 +00:00
Dag-Erling Smørgrav 9cb460b85a Simplify the log accumulation code. It turns out that the trick of
looking for ^===> can give quite annoying false positives, especially
when building kernels, so drop it; the context can be inferred from
make's "Stop in /foo/bar/baz" messages anyway.

Also add a case that I'd missed the first time around (which happens
to be the common case, not the exception...)
2003-02-12 21:06:19 +00:00
Dag-Erling Smørgrav bd75cb6142 NO_KERNELCLEAN was a booboo. 2003-02-12 20:29:05 +00:00
Dag-Erling Smørgrav 42d534b255 Welcome tbmaster, TinderboxNG's better half. 2003-02-12 20:23:27 +00:00
Dag-Erling Smørgrav c2b94638c3 Slight cleanup; adjust verbosity; unlink the logfile before opening it
to tickle 'tail -F'.
2003-02-12 20:18:08 +00:00
Dag-Erling Smørgrav a2036ac045 Teach whereintheworld to preserve TinderboxNG log messages. 2003-02-12 09:51:47 +00:00
Dag-Erling Smørgrav 487106ce7a Make "CURRENT" the default branch, and special-case it in the update
code so it results in -A rather than -rCURRENT.
2003-02-12 09:13:36 +00:00
Dag-Erling Smørgrav 55b84e8a54 Say hello to Tinderbox 2.0, the choice of a new generation! 2003-02-11 22:17:49 +00:00
Garance A Drosehn 413a16d889 The regression-tests for 'make' *expect* to trigger make's warning:
warning: duplicate script for target "double" ignored
The regression-tests do try to hide that message, but the message does
still appear when using -j (eg: 'make -j5 buildworld').  This changes the
regression-test so the expected warning message will not be seen even
when -j is specified.

Reviewed by:	jmallett  ru
2003-02-10 19:56:42 +00:00
Mike Barcroft b4e41c1278 Add more tests. Specifically, infinity cast from other types should
always be classified as infinity.
2003-02-08 20:45:48 +00:00
David Xu 10fece7726 Add a missing inline keyword. 2003-02-08 06:33:37 +00:00
Mike Barcroft 2c4594131f Add regression test for soon-to-be-committed fpclassify(). 2003-02-05 16:40:18 +00:00