[00:06:41] <CIA-110> kallisti5 * r41722 /haiku/trunk/ (3 files in 3 dirs): (log message trimmed)
[00:06:41] <CIA-110> * rename graphics_memory to frame_buffer. lets keep consistant
[00:06:41] <CIA-110> * pass mapped frame buffer area id to accelerant
[00:06:41] <CIA-110> * remove my temporary hacked together frame buffer memory mapping
[00:06:41] <CIA-110> * completely rely on PCI BAR for now for aperture size / location instead of
[00:06:42] <CIA-110> R6XX_CONFIG_FB_BASE reg.
[00:06:42] <CIA-110> * Remove my temporary AllocateFB function.
[01:28:25] <CIA-110> bonefish * r41723 /haiku/branches/developer/bonefish/signals/ (3 files in 2 dirs):
[01:28:26] <CIA-110> Solved setjmp()/longjmp() binary compatibility issue: The jmp_buf structure
[01:28:26] <CIA-110> contains a sigset_t field which has changed in size. By storing the inverted
[01:28:27] <CIA-110> mask we can drop the mask_was_saved field, so that the structure still has the
[01:28:27] <CIA-110> same size and nothing more needs to be done to maintain binary compatibility.
[01:28:28] <CIA-110> Fixes issues with the nano package, BTW.
[01:37:36] <CIA-110> bonefish * r41724 /haiku/branches/developer/bonefish/signals/src/system/libroot/posix/sys/select.c:
[01:37:36] <CIA-110> pselect() has sigset_t* parameter. Added versioned symbols for binary
[01:37:36] <CIA-110> compatibility.
[01:41:27] <CIA-110> bonefish * r41725 /haiku/branches/developer/bonefish/signals/src/system/kernel/kernel_versions: Added KERNEL_1_ALPHA4 version node, needed for pselect().
[02:09:19] <CIA-110> mmlr * r41726 /haiku/trunk/src/system/kernel/arch/x86/irq_routing_table.cpp: (log message trimmed)
[02:09:19] <CIA-110> * Turn around finding the routing table on a PCI bridge and checking for it to
[02:09:19] <CIA-110> actually be a bridge. We always want to check if the device at hand is a PCI
[02:09:19] <CIA-110> bridge, because in that case we need to recurse further down. We recurse now
[02:09:19] <CIA-110> even if there is no routing table found for the current bridge, since that
[02:09:19] <CIA-110> doesn't mean that another bridge, further below, can't have a routing table
[02:09:20] <CIA-110> again. Possibly fixes #7520.
[02:39:42] <CIA-110> bonefish * r41727 /haiku/trunk/src/tests/system/libroot/posix/posixtestsuite/conformance/interfaces/sigset/6-1.c:
[02:39:42] <CIA-110> Fixed check. sigset() is supposed to return the previous signal disposition,
[02:39:42] <CIA-110> respectively SIG_HOLD, if the signal was blocked.
[02:43:56] <CIA-110> bonefish * r41728 /haiku/trunk/src/tests/system/libroot/posix/posixtestsuite/conformance/interfaces/sigset/6-1.c: Reverted r41727. Accidentally committed to wrong branch.
[02:45:01] <CIA-110> bonefish * r41729 /haiku/branches/developer/bonefish/signals/src/tests/system/libroot/posix/posixtestsuite/conformance/interfaces/sigset/6-1.c:
[02:45:01] <CIA-110> Fixed check. sigset() is supposed to return the previous signal disposition,
[02:45:01] <CIA-110> respectively SIG_HOLD, if the signal was blocked.
[02:58:43] <CIA-110> bonefish * r41730 /haiku/branches/developer/bonefish/signals/src/tests/system/libroot/posix/posixtestsuite/conformance/interfaces/sigset/7-1.c: Fixed check: sigset() returns the previous signal disposition.
[02:59:14] <CIA-110> bonefish * r41731 /haiku/branches/developer/bonefish/signals/src/tests/system/libroot/posix/posixtestsuite/conformance/interfaces/sigset/8-1.c: Changed broken test into something useful.
[03:55:32] <CIA-110> bonefish * r41732 /haiku/branches/developer/bonefish/signals/src/tests/system/libroot/posix/posixtestsuite/conformance/interfaces/sigwait/3-1.c:
[03:55:32] <CIA-110> Nice test, but unfortunately it is implementation-defined whether in cases
[03:55:32] <CIA-110> where signal queuing is not required signals are queued nonetheless. Haiku
[03:55:32] <CIA-110> queues agressively and therefore there are still 3 of the 4 signals pending
[03:55:32] <CIA-110> after sigwait() returns. Removed the test.
[03:57:57] <CIA-110> bonefish * r41733 /haiku/branches/developer/bonefish/signals/src/tests/system/libroot/posix/posixtestsuite/conformance/interfaces/sigwait/ (assertions.xml coverage.txt): Removed the assertion as well.
[04:08:12] <CIA-110> bonefish * r41734 /haiku/branches/developer/bonefish/signals/src/system/libroot/posix/signal/ (sigtimedwait.cpp sigwait.cpp):
[04:08:12] <CIA-110> Fixed sigwait() and sigtimedwait(). sigwait() returns 0 on success and the error
[04:08:12] <CIA-110> code on failure (no errno). sigtimedwait() returns the signal number on success
[04:08:12] <CIA-110> and -1 on failure (setting errno).
[04:10:59] <CIA-110> kallisti5 * r41735 /haiku/trunk/build/scripts/build_haiku_image:
[04:11:00] <CIA-110> add an attempted fallback to genisoimage if mkisofs doesn't exist on the system.
[04:11:00] <CIA-110> (debian no longer ships mkisofs) genisoimage is based on an older version of
[04:11:00] <CIA-110> mkisofs and has issues with long filenames. The genisoimage iso-level override
[04:11:00] <CIA-110> makes the image work and allows long filenames but may introduce issues? placed
[04:11:00] <CIA-110> a warning to the user that genisoimage is a fallback
[04:54:32] <CIA-110> anevilyak * r41736 /haiku/trunk/src/system/ (boot/loader/menu.cpp kernel/debug/blue_screen.cpp):
[04:54:32] <CIA-110> Add debug option to disable the on-screen debug output's paging by default, for
[04:54:32] <CIA-110> cases such as USB keyboards where the keyboard no longer functions once the
[04:54:32] <CIA-110> kernel is entered and the USB stack initializes.
[04:54:33] <CIA-110> Implements #7561.
[08:43:55] *** kurain has joined #haiku-jp
[10:47:52] *** kurain has quit IRC
[10:52:04] *** kurain has joined #haiku-jp
[11:05:05] *** kurain has quit IRC
[11:05:30] *** kurain has joined #haiku-jp
[11:11:32] *** kurain has quit IRC
[11:13:16] *** kurain has joined #haiku-jp
[15:03:27] <CIA-110> scottmc * r41737 /haiku/trunk/build/jam/OptionalPackages:
[15:03:27] <CIA-110> Updated gcc4 taglib package and fixed typo in mercurial's filename.
[15:03:27] <CIA-110> +alpha3
[17:05:24] *** kurain has quit IRC
[17:05:58] <CIA-110> bonefish * r41738 /haiku/branches/developer/bonefish/signals/src/system/kernel/thread.cpp:
[17:05:58] <CIA-110> Added KernelThreadCreationAttributes::signal_mask. create_thread() sets the new
[17:05:58] <CIA-110> thread's sig_block_mask to that value. Used by _user_spawn_thread() to let the
[17:05:58] <CIA-110> new thread inherit the calling thread's signal mask as required by POSIX.
[17:17:17] <CIA-110> bonefish * r41739 /haiku/branches/developer/bonefish/signals/src/system/libroot/posix/signal/set_signal_mask.cpp:
[17:17:17] <CIA-110> Fixed return values: pthread_setmask() returns the error code drectly,
[17:17:58] <CIA-110> sigprocmask() sets errno instead.
[17:18:25] <CIA-110> bonefish * r41740 /haiku/branches/developer/bonefish/signals/src/libs/posix_error_mapper/signal.cpp: Added wrapper for sigwait(), since it returns the error code directly.
[17:46:18] <CIA-110> axeld * r41741 /haiku/trunk/src/kits/locale/Country.cpp: * Minor cleanup.
[18:01:12] <CIA-110> bonefish * r41742 /haiku/branches/developer/bonefish/signals/src/system/kernel/signal.cpp:
[18:01:13] <CIA-110> sigwait_internal(): Map B_TIMED_OUT to B_WOULD_BLOCK (EAGAIN), which is the
[18:01:13] <CIA-110> error code demanded by POSIX.
[18:02:12] <CIA-110> bonefish * r41743 /haiku/branches/developer/bonefish/signals/src/system/libroot/posix/signal/raise.c: raise() must set errno instead of returning the error code directly.
[19:07:16] <CIA-110> axeld * r41744 /haiku/trunk/ (4 files in 2 dirs):
[19:07:16] <CIA-110> * Added BLanguage::GetIcon(), and BLocaleRoster::GetFlagIconForLanguage(). The
[19:07:16] <CIA-110> former just calls the latter. Getting the flag for a language is pretty
[19:07:16] <CIA-110> simplistic for now, but it won't return the wrong flag, just only a few known
[19:07:16] <CIA-110> ones (should be enough for ReadOnlyBootPrompt, at least).
[19:07:16] <CIA-110> * Ordered methods in declaration order.
[19:08:11] <CIA-110> axeld * r41745 /haiku/trunk/src/apps/readonlybootprompt/BootPromptWindow.cpp:
[19:08:11] <CIA-110> * Reenabled flags via the new BLocaleRoster::GetFlagIconForLanguage() method.
[19:08:11] <CIA-110> +alpha (also for the previous commit)
[19:29:01] <CIA-110> bonefish * r41746 /haiku/branches/developer/bonefish/signals/ (9 files in 5 dirs): (log message trimmed)
[19:29:01] <CIA-110> Removed the TLS slot for the thread's pthread_t (TLS_PTHREAD_SLOT). Instead
[19:29:01] <CIA-110> added a pthread field to the user_thread structure. The pointer is passed into
[19:29:01] <CIA-110> the kernel via a new thread_creation_attributes field (also stored in the new
[19:29:01] <CIA-110> Thread::pthread field) and set in the user_thread structure by the kernel. This
[19:29:02] <CIA-110> has the advantage that the pthread_t is associated with the thread even before
[19:29:03] <CIA-110> it starts execution at its entry function (which would set the TLS entry). This
[20:06:51] <CIA-110> bonefish * r41747 /haiku/branches/developer/bonefish/signals/src/tests/system/kernel/unit/kernel_unit_tests.cpp: Fixed build.
[20:19:37] <CIA-110> pulkomandy * r41748 /haiku/branches/releases/r1alpha3/ (29 files in 22 dirs): Merge 41711,41720,41737
[20:22:06] <CIA-110> bonefish * r41749 /haiku/branches/developer/bonefish/signals/src/system/kernel/TeamThreadTables.h:
[20:22:06] <CIA-110> TeamThreadTable::NextElement(): Insert the iterator entry after the found entry
[20:22:06] <CIA-110> as the comment says. Fixes infinite loop when actually using a
[20:22:06] <CIA-110> {Team,Thread}ListIterator.
[20:23:19] <CIA-110> bonefish * r41750 /haiku/branches/developer/bonefish/signals/src/system/kernel/debug/system_profiler.cpp:
[20:23:19] <CIA-110> SystemProfiler::Init(): Fixed inverse abort check in the team and the thread
[20:23:19] <CIA-110> iteration loops.
[22:23:17] <CIA-110> bonefish * r41751 /haiku/branches/developer/bonefish/signals/ (4 files in 3 dirs): Moved send_signal() and set_signal_stack() from <signal.h> to <OS.h>.





top