src - OpenBSD base system

Age	Commit message (Collapse)	Author
2017-08-04	Use C stdio instead of C++ iostream as we cannot use the latter in	Mark Kettenis
	constructors for an "initfirst" object as the iostream objects have not been constructed yet. ok bluhm@
2017-08-04	Use ${CC}=${CXX} here as well.	Mark Kettenis
	ok bluhm@
2017-08-04	Use volatile to keep clang from optimizing things too much. Force linking	Mark Kettenis
	with the c++ command instead of cc by setting CC=${CXX}. This is a hack and a better solution is under development. Linking with c++ is necessary to make the shared libraries depend up libc++.so such that the iostream objects are constructed before we use them. ok bluhm@
2017-08-02	Split up source code to keep clang from optimizing this into failure.	Mark Kettenis
	ok millert@
2005-09-28	two new regresses for dlclose. double dlopen("foo") with single	Kurt Miller
	dlclose(foo). mozilla lazy binding example (load_object confusion)
2017-08-01	A fix is coming in the future, but in the meantime we should never declare	Theo de Raadt
	a section(".openbsd.randomdata") variable as static or const, because the compiler optimizes out access.
2017-07-31	Tweak demoted state test, ok jca@	rob

2017-07-31	Fix whitespace, ok jca@	rob

2017-07-30	Add shebang	Jeremie Courreges-Anglas

2017-07-29	Remove unused function and variable to shut up warning. Put back	Alexander Bluhm
	a status print that was lost in previous.
2017-07-29	repair regress for more aggressive post fork checking in library.	Ted Unangst
	we should not be checking that things explicitly forbidden work. prompted by bluhm
2017-07-29	Use linker flags that build and run the test with gcc and clang.	Alexander Bluhm
	Add missing #include to fix clang warning.
2017-07-28	fix date	Ingo Schwarze

2017-07-28	Do not test .Xr here. It is not what this directory should focus on,	Ingo Schwarze
	and it is causing too much trouble in -portable.
2017-07-28	Don't call fatal from stop_sshd since it calls cleanup which calls stop_sshd	Darren Tucker
	which will probably fail in the same way. Instead, just bail. Differentiate between sshd dying without cleanup and not shutting down.
2017-07-27	Sort subdir, run getaddrinfo test, remove vax case.	Alexander Bluhm

2017-07-27	Sort expected getaddrinfo output in inet4 inet6 order, then the	Alexander Bluhm
	test passes.
2017-07-27	Fix all clang warnings in libc regress and cleanup some make files.	Alexander Bluhm

2017-07-27	Compile and test MPATH code.	Martin Pieuchot

2017-07-27	Work around for rttest8 that some times fail due to a reference count	Martin Pieuchot
	race in the kernel.
2017-07-27	Mark the invalid memory location as volatile, otherwise clang would	Alexander Bluhm
	optimize the access to it with an illegal instruction. But the tests needs a SIGSEGV, it would fail with SIGILL.
2017-07-27	fcntl(F_DUPFD) needs an argument. Otherwise it fails with EINVAL	Alexander Bluhm
	when compiled with clang.
2017-07-25	Remove dynamically created files from CVS tree.	Alexander Bluhm

2017-07-25	Do not remove files from CVS tree during make clean.	Alexander Bluhm

2017-07-24	Extend the SNI tests to ensure that we have a hostname in the session	Joel Sing
	after a successful tlsext_sni_serverhello_parse() and that tlsext_sni_clienthello_parse() fails if we have an existing session and the SNI is mismatched.
2017-07-24	Add regress coverage for the TLS Renegotiation Indication extension.	Joel Sing

2017-07-22	Tolerate imprecise fault address on sparc64 where the hardware doesn't provide	Mark Kettenis
	tho low order bits. ok deraadt@, visa@
2017-07-22	Add missing RCS IDs and zap redundant SRCS from Makefile.	anton

2017-07-20	extra test because mk38 triggers two issues and fixing the first one	Marc Espie
	makes the second one vanish
2017-07-20	Fixup pf109.ok and rewrite pf110	Mike Belopuhov

2017-07-20	Accessing a mmap(2)ed file behind its end should result in a SIGBUS	Alexander Bluhm
	according to POSIX. Bring regression test and kernel in line for amd64 and i386. Other architectures have to follow. OK deraadt@ kettenis@
2017-07-20	Avoid cross reference to mandoc(1) because it may cause a message	Ingo Schwarze
	on other operating systems where mandoc is not installed, causing the test to fail in the portable test suite.
2017-07-19	The first version of the edit program used to test csh, ksh and mail was based	anton
	on the following reasoning: once the program to test has written some data it has entered the main-loop and is by now ready to receive user input. At this point it should be safe to start writing input and once the program once again enters a blocking reading state, its done processing the input. This approach was sensitive to timing and determining when a shell is done processing its input (if ever) is tricky. This iteration of the edit program takes a new approach and uses presence of a prompt for synchronisation of I/O. It doesn't solve all problems but is a step in the right direction. Joint work with bluhm@
2017-07-19	Fix failing tests when executed as a user without a ~/.mailrc configuration ↵	anton
	file. While here, fix a flaky test.
2017-07-19	Regression tests asserting various HFSC and FQ-CoDel functionality	Mike Belopuhov

2017-07-18	The function getaddr() compares the stack address of the caller	Alexander Bluhm
	with its own local variable. This way it detects the direction of stack growth. This does not work if the functions gets inlined. Pass -fno-inline to the compiler to avoid sporadic tests fails.
2017-07-16	Provide a new regress test for TLS extension handlers, currently covering	Joel Sing
	the newly converted SNI code.
2017-07-16	fix dates	Ingo Schwarze

2017-07-16	test -diag -width and -inset -width	Ingo Schwarze

2017-07-15	Add a test that covers the recently fixed "0x" prefix handling issue.	Joel Sing

2017-07-15	Catch up with eqn.c rev. 1.41 by bentley@:	Ingo Schwarze
	Print the right character for mathematical overbar.
2017-07-15	regress/sbin/pfctl - use -n for pfcmd group	Adam Wolk
	Passing in the -n which just results in the rules being parsed. The definition for the pfctlcmd testing group in the Makefile is: # pfcmd: test pfctl command line parsing This test passeed only because pfctl didn't exit with an error when it was asked to clear tables from a non existing anchor. with the -n flag we still test if the anchor command can be combined with -Fa but don't actually attempt to run the clearing code. OK mikeb@, sashan@
2017-07-14	adapt to hex format of character entities,	Ingo Schwarze
	committed by bentley@ in html.c rev. 1.86
2017-07-14	Do not mix Perl read() with sysread(). Data could get stuck in the	Alexander Bluhm
	buffered IO and test run-regress-args-http-chunked.pl would fail on slow hardware. Limit debug output to one line per 1% of data. Improve function write_syswrite(). Sync with httpd regress. Fix whitespace.
2017-07-14	Do not mix Perl read() with sysread(). Data could get stuck in the	Alexander Bluhm
	buffered IO and test run-regress-args-get-1073741824.pl would fail on slow hardware. Introduce a common function read_part() that uses Perl read(). Limit debug output to one line per 1% of data. Remove unused function http_server(). Fix whitespace. Cleanup Makefile.
2017-07-14	Add regression tests for ifstated.	Sebastian Benoit
	Not hooked into regress/usr.sbin/Makefile yet. From Rob Pierce <rob@2keys.ca>
2017-07-13	Round VM_MAXUSER_ADDRESS down to next page to make mmap_hint regress	Alexander Bluhm
	pass on sparc64.
2017-07-13	Improve error messages of siginfo-fault regress to make clear which	Alexander Bluhm
	tests fail. First run all checks and report problmes, finally exit with the total number of failures. Unfortunately it still fails.
2017-07-12	We cannot mmap(2) memory at VM_MAXUSER_ADDRESS as the full allocated	Alexander Bluhm
	page must be below that address. Subtract PAGE_SIZE from the hint address, then regress passes.
2017-07-12	Disable failing pkg_add tests and document when they were added.	Alexander Bluhm
	Targets are still executed, but are expected to fail.