time. Releasing the cache's store reference while holding the cache lock
could reverse the usual locking order -- the VFS could potentially call
the remove_vnode() or put_vnode() FS hook, which in turn could use the
file cache, thus resulting in a deadlock. Now we release the store ref
before locking the cache.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24548 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Install libncurses.a.
* Copy the cpp headers from our repository; don't use the compiler
headers. This needs some more shuffling around.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24543 a95241bf-73f2-0310-859d-f6bbb57e9c96
gcc. You need to reconfigure and rebuild the cross-tools to have this
change take effect.
Note that from now on it is no longer possible to use the native BeOS
compiler to build Haiku. You'll have to build a cross compiler, too. I
haven't tested whether this works at all, though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24542 a95241bf-73f2-0310-859d-f6bbb57e9c96
depending on the gcc version of the executable.
* Adjusted non-BeOS-style symbol resolution so that add-ons and
dynamically loaded libraries find symbols in the executable.
This change re-enables support for undefined symbols.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24537 a95241bf-73f2-0310-859d-f6bbb57e9c96
shared object to find out the GCC version it has been compiled with.
This is not an exact science, since we've got the version string for
every single object file that has been linked in, but my heuristic seems
to be good enough.
Having the gcc version at hand will allow for two features: Enabling
work-arounds for old executables (like the type info problem in BeOS
apps), and automatically selecting the right set of libraries in a mixed
gcc 2/gcc 4 environment.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24534 a95241bf-73f2-0310-859d-f6bbb57e9c96
so one can set it to the static libroot, if desired.
* Generic attribute emulation:
- Added build tool rm_attrs, a simple "rm" replacement, which also
removes the attributes directory for a given file.
- Added build/scripts/rm_attrs shell script, which wraps the
invocation of the rm_attrs tool. If it doesn't exist yet, the
ordinary rm is used.
- The RM jam variable refers to the rm_attrs script now, i.e. whenever
something is removed by the build system, the attributes are removed
too (if the build tool has already been built, that is).
- Removed the shell function attrrmrf() in build_haiku_image. We use
the rm_attrs tool instead, if necessary.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24528 a95241bf-73f2-0310-859d-f6bbb57e9c96
which returns the attribute directory for a given file's stat.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24527 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Also install _G_config.h.
* /boot/develop/headers/cpp is a symlink to g++'s headers, now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24523 a95241bf-73f2-0310-859d-f6bbb57e9c96
executable permissions of the file by hand. We use _kern_access()
instead, which also handles the root case correctly.
* The user and group arguments of test_executable() aren't needed any
longer.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24483 a95241bf-73f2-0310-859d-f6bbb57e9c96
and returns int. Fixes the only (unexpectedly) failing autoconf test.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24480 a95241bf-73f2-0310-859d-f6bbb57e9c96
prevent "configure" scripts from thinking we do actually support it.
Having real-time signals would be nice though (cf. #1935).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24444 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Build generate_boot_screen as a build platform tool. It requires
libpng to be installed.
* Added output file name as fourth argument.
* Replaced unnecessarily complicated
fwrite(<string>, strlen(<string>),...) and sprintf()+fwrite()
constructs by fputs() and fprintf() respectively.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24440 a95241bf-73f2-0310-859d-f6bbb57e9c96
picked might have changed while we were locking its cache. Might fix
#1931.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24430 a95241bf-73f2-0310-859d-f6bbb57e9c96
readlink() function. It is no longer required to null-terminate the
string, shall not fail, if the buffer is too small, and shall return
the length of the string actually written into the buffer.
* Adjusted rootfs, devfs, and bfs accordingly. Also adjusted their
read_stat() hooks to return the correct symlink length in st_size.
* Our readlink() does now comply to the standard (and BeOS).
Additionally if the buffer is big enough it is nice to non-conforming
apps and null-terminates it.
* BSymLink::ReadLink() explicitly null-terminates the string now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24425 a95241bf-73f2-0310-859d-f6bbb57e9c96
builds on Linux with gcc 2 and 4, but the images do even run. :-) Not
tested on BeOS.
* Removed stddef.h and stdarg.h. They are provided by the compiler.
* Adjusted size_t.h, wchar_t.h, and wchar.h accordingly.
* Made stdio.h avoid gcc 2.95.3's fixincludes hack stdio_va_list
* Added gcc 2.95.3 headers to the repository. They are used instead of
the headers of the gcc 2.95.3's we use to compile Haiku with. Should
avoid build problems with the BeOS native compiler.
For sake of personal recreation you can rebuild the cross gcc 2.95.3,
but the only thing that changed is its header directory
(lib/gcc-lib/.../include), which isn't used anymore. Replacing it with
headers/build/gcc-2.95.3 should have the same effect as rebuilding, BTW.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24413 a95241bf-73f2-0310-859d-f6bbb57e9c96
hack math_huge_val_ifndef does, anyway. We do it ourselves and remove
the therefore superfluous gcc math.h header.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24412 a95241bf-73f2-0310-859d-f6bbb57e9c96
change resulted in "version mismatch between boot loader and kernel". So
apparently the size of some type changed unintentionally.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24408 a95241bf-73f2-0310-859d-f6bbb57e9c96
under Haiku overrides ours anyway.
* Adjusted size_t.h, wchar_t.h, wchar.h accordingly. This should fix the
annoying "ssize_t redefined" warnings when compiling under Haiku.
* When building Haiku the gcc headers come first in the include
search path now, as it should be. The respective TODO suggested that
this might break the build depending on compiler version and host
platform. I've tested with Linux gcc 2 and gcc 4, which work fine.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24406 a95241bf-73f2-0310-859d-f6bbb57e9c96
number. Was a problem only for partitions > 2^32 * block size (4TB
for 1KB blocks).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24393 a95241bf-73f2-0310-859d-f6bbb57e9c96
from the call to vfs_get_vnode() now. Only this way it is safe to call
store_release_ref() later (as the page writer does). We had a potential
race condition -- if called after vm_cache_remove_consumer() had
released the last reference, the old vnode might already have been
deleted.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24386 a95241bf-73f2-0310-859d-f6bbb57e9c96
thread_yield() calls make reseed() really slow in a busy system, since
another thread will exhausts its full quantum between two thread_yield()
calls. And 257 * 8 * 8 quanta are about 49s.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24375 a95241bf-73f2-0310-859d-f6bbb57e9c96
%edx) was trashed before passing it to the user debugger. strace shows
the correct return values again.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24364 a95241bf-73f2-0310-859d-f6bbb57e9c96
only gets the caller into trouble, since then another signal should be
sent.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24363 a95241bf-73f2-0310-859d-f6bbb57e9c96
structure. They are properly inherited and updated on
fork(), load_image(), and exec().
* Implemented the get[e]{u,g}id(), set[[r]e]{u,g}id() family for real.
* getgroups() also calls the kernel now, but only returns the effective
group ID. Supplementary groups support is still missing.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24359 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Got rid of class ObjFunctionSummand. Both the constraint summands and
the objective function summands are now stored using class Summand.
* Some method names are more BeOS compliant now: SetX instead of ChangeX.
* linprog test code now uses new AddConstraint methods.
* CalculateMinSize and CalculateMaxSize did not free the memory they
allocated.
* Removed inappropriate setter and getter methods.
* Memory allocated in class Constraint is freed now.
* Other small changes.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24351 a95241bf-73f2-0310-859d-f6bbb57e9c96
uncommented the meaty lines again. Added one or two more explaining
sentences.
* Created a concise UserBuildConfig.sample with very few comments and
only the most interesting (commented out) example lines for those
people who don't read the UserBuildConfig.ReadMe.
I hope everyone can live with this solution.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24350 a95241bf-73f2-0310-859d-f6bbb57e9c96
advantages that is supposed to yield, but at least under Linux that
option is enabled.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24343 a95241bf-73f2-0310-859d-f6bbb57e9c96
all FDs in a fork()+exec()ed child open. Makes Perl test
t/run/cloexec.t pass.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24336 a95241bf-73f2-0310-859d-f6bbb57e9c96
as the death entries of a deceased processes that were in this group
and have not yet been reaped hold references to the group, so that it
won't be deleted until the group is empty and all gone group members
have been reaped. This fixes#1799 at last.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24335 a95241bf-73f2-0310-859d-f6bbb57e9c96
programs happy, that do that unconditionally (e.g. some Perl test).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24334 a95241bf-73f2-0310-859d-f6bbb57e9c96
in code that has interrupts disabled. The chunks of memories are queued
and free()d periodically by a kernel daemon.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24332 a95241bf-73f2-0310-859d-f6bbb57e9c96
truncated, if already existing. This problem could be experienced when
saving in Pe.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24329 a95241bf-73f2-0310-859d-f6bbb57e9c96
I tested building libroot.so only (BeOS under VMware is not exactly
fun as a build platform), but I think that should suffice.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24134 a95241bf-73f2-0310-859d-f6bbb57e9c96
* copyright headers for the files of the libraries linprog and alm
* new class Summand for representing summands in a linear constraint
* merged class SoftConstraint into class Constraint; Constraint now
supports both soft and hard constraint functionality
* new AddConstraint methods in class LinearSpec for directly setting
constraints with 1 to 4 summands
* code cleanups by using aforementioned AddConstraint methods
* a new very simple test application for alm
* some style corrections
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24108 a95241bf-73f2-0310-859d-f6bbb57e9c96
Fixed building Haiku images on FreeBSD, which was broken since the
introduction of the device support.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24107 a95241bf-73f2-0310-859d-f6bbb57e9c96
headers/os version and removed the now duplicate icon type constant
definitions from several sources.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24106 a95241bf-73f2-0310-859d-f6bbb57e9c96
need to make sure that we never get that far with a negative timeout or
we'll wait forever.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24083 a95241bf-73f2-0310-859d-f6bbb57e9c96
timeout, not just -1. Paranoia only, since that should happen anyway.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24082 a95241bf-73f2-0310-859d-f6bbb57e9c96
negative relative timeout would be converted to B_INFINITE_TIMEOUT.
For some reason the ScreenSaver preflet is snooze()ing with such
a value.
* Some more comments.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24080 a95241bf-73f2-0310-859d-f6bbb57e9c96
interrupted early, i.e. when signals were already pending when we
started waiting. Fixes#1832.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24063 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added team::flags. Currently only used for setting a flag when a team
has exec()ed.
* Some improvements of _user_setpgid():
- It failed incorrectly when the target process was a process group
leader. According to the standard it shall fail when the process is
a session leader. Moving a process group leader to another process
group is fine, even if that leaves the group leaderless.
- Fixed race conditions. We need to recheck the error conditions when
we hold the team spinlock. Otherwise the situation could change
while we allocated the new process group. This was one of the
reasons for bug #1799 -- after the shell fork()'s both parent and
child invoke setpgid() for the child.
- Fixed behavior for pid == pgid. It doesn't necessarily mean that a
new group has to be created.
- Fixed update of target process group orphaned state.
- Squashed TODO: setpgid() on a child is supposed to fail after the
child has exec()ed.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24041 a95241bf-73f2-0310-859d-f6bbb57e9c96
success.
* setpgrp() is not supposed to fail (could happen, if the calling
process was a session leader).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24040 a95241bf-73f2-0310-859d-f6bbb57e9c96
allocated on the stack, condition variable related structures would be
trashed, causing all kinds of problems. Fixes#1811 and #1812.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24009 a95241bf-73f2-0310-859d-f6bbb57e9c96
but there's a special handling for 0 us relative timeouts. Syscalls
usually return B_WOULD_BLOCK instead of B_TIMED_OUT in this case, and
callers might explicitely check for it. Hence we don't convert 0 us
timeouts anymore. gdb works again.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23989 a95241bf-73f2-0310-859d-f6bbb57e9c96
Nested syscall restarts (interrupted syscall in a signal handler)
aren't tested yet.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23985 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added syscall restart support for connect(), accept(), send(), recv(),
which are implemented via ioctl()s. The actual restart support is done
in the net stack driver's ioctl() hook. Lower layers need to correctly
deal with socket timeouts, though, for which the stack module provides
support functions.
* TCPEndpoint::_WaitForEstablished() does abort now when an error
occurred earlier, so that trying to connect to an unused port fails
immediately, as it should.
* Fixed and refactored TCP connection reset handling. The new
TCPEndpoint::_HandleReset() does the job. Got rid of
TCPEndpoint::fError.
* Fixed sequence numbers for SYNC/FINI packets.
* The former two fix the problem that connections wouldn't be closed
correctly and could even be reused when trying to connect again (as
was reproducible with svnserve + svn).
* Some style cleanup in CPEndpoint.h.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23984 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Implemented automatic syscall restarts:
- A syscall can indicate that it has been interrupted and can be
restarted by setting a respective bit in thread::flags. It can
store parameters it wants to be preserved for the restart in
thread::syscall_restart::parameters. Another thread::flags bit
indicates whether it has been restarted.
- handle_signals() clears the restart flag, if the handled signal
has a handler function installed and SA_RESTART is not set. Another
thread flag (THREAD_FLAGS_DONT_RESTART_SYSCALL) can prevent syscalls
from being restarted, even if they could be (not used yet, but we
might want to use it in resume_thread(), so that we stay
behaviorally compatible with BeOS).
- The architecture specific syscall handler restarts the syscall, if
the restart flag is set. Implemented for x86 only.
- Added some support functions in the private <syscall_restart.h> to
simplify the syscall restart code in the syscalls.
- Adjusted all syscalls that can potentially be restarted accordingly.
- _user_ioctl() sets new thread flag THREAD_FLAGS_IOCTL_SYSCALL while
calling the underlying FS's/driver's hook, so that syscall restarts
can also be supported there.
* thread_at_kernel_exit() invokes handle_signals() in a loop now, as
long as the latter indicates that the thread shall be suspended, so
that after waking up signals received in the meantime will be handled
before the thread returns to userland. Adjusted handle_signals()
accordingly -- when encountering a suspending signal we don't check
for further signals.
* Fixed sigsuspend(): Suspending the thread and rescheduling doesn't
result in the correct behavior. Instead we employ a temporary
condition variable and interruptably wait on it. The POSIX test
suite test passes, now.
* Made the switch_sem[_etc]() behavior on interruption consistent.
Depending on when the signal arrived (before the call or when already
waiting) the first semaphore would or wouldn't be released. Now we
consistently release it.
* Refactored _user_{read,write}[v]() syscalls. Use a common function for
either pair. The iovec version doesn't fail anymore, if anything could
be read/written at all. It also checks whether a complete vector
could be read/written, so that we won't skip data, if the underlying
FS/driver couldn't read/write more ATM.
* Some refactoring in the x86 syscall handler: The int 99 and sysenter
handlers use a common subroutine to avoid code duplication.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23983 a95241bf-73f2-0310-859d-f6bbb57e9c96
Before starting to wait on a condition variable check for pending
signals first, if the call is interruptable.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23980 a95241bf-73f2-0310-859d-f6bbb57e9c96
The return value of Inode::WaitForRequest() is status_t not bool. So the
method would always fail when it actually succeeded. This affected reads
from pipes which didn't have data. The bug was hidded since VFS code
mostly checks error codes only against < B_OK, so that such a read would
be treated as 0 byte read.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23976 a95241bf-73f2-0310-859d-f6bbb57e9c96
BeOS R5's glue code incorrectly calls _thread_do_exit_notification()
when main() returns, while Haiku does that in exit(). Therefore when
terminating this way the exit hooks were called twice for executables
built under BeOS R5. This caused e.g. NetPositive or the R5 svn to
crash on exit (our network code actually uses those hooks).
Fixes bug #1742.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23975 a95241bf-73f2-0310-859d-f6bbb57e9c96
Adjusted PATH, LIBRARY_PATH, and the paths built into the runtime loader
to included the /boot/common tree.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23974 a95241bf-73f2-0310-859d-f6bbb57e9c96
Removed the "config" subdirectory in the /boot/common directory
structure.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23973 a95241bf-73f2-0310-859d-f6bbb57e9c96
Changed the way the VMware image is built. Instead of creating a normal
image first and then adding vmdk header and that image, we create the
VMware image in one go, now. Therefore "jam update-vmware-image ..."
does now actually update the VMware image directly, instead of updating
the normal image and recreating the VMware image.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23972 a95241bf-73f2-0310-859d-f6bbb57e9c96
Added parameter --start-offset to allow writing the boot code not only
at the beginning of the given file.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23971 a95241bf-73f2-0310-859d-f6bbb57e9c96
Added parameters --start-offset and --end-offset to restrict the
access of the file system to only that part of the given file.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23970 a95241bf-73f2-0310-859d-f6bbb57e9c96
some absolute path was not enough to always recognize a library as
already loaded. This fixes problems with Perl where loading an add-on
would cause another instance of libperl.so to be loaded, which would
lead to crashes due to uninitialized static vars in the new instance.
Perl builds now and the tests run, but quite a few do fail.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23930 a95241bf-73f2-0310-859d-f6bbb57e9c96
using the int 99 syscall method. Otherwise it would remain set e.g.
after _kern_restore_signal_frame() and the next syscall would look like
one returning a 64 bit value.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23918 a95241bf-73f2-0310-859d-f6bbb57e9c96
function has the old behavior. When false, it just calls the scheduler
without any priority adjustment or other stuff.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23906 a95241bf-73f2-0310-859d-f6bbb57e9c96
/dev/urandom. It verifies that Haiku's dev/urandom is performance-wise
totally unusable -- it takes several seconds.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23904 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added temporary copyright entry for lp_solve as a placeholder, until I
know what it should say.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23890 a95241bf-73f2-0310-859d-f6bbb57e9c96
Ingo Weinhold