Commit by mmlr on master :: rhrev43823 haiku/: (link)

Add scripting property for toggling fullscreen.

Commit by mmlr on master :: rhrev43717 haiku/: (link)

Update wpa_supplicant package for gcc2.

It should work on current images again and contains a few usability
tweaks that make it simpler to navigate/use.

Commit by mmlr on master :: rhrev43638 haiku/: (link)

Remove coreutils and tcpdump from the DEBUG blacklist. [2 commits]

The DEBUG build for these has been fixed in hrev43636 and hrev43637
respectively.

Commit by mmlr on master :: rhrev43637 haiku/: (link)

Use static inline instead of extern inline to fix DEBUG build.

Fixing this directly even though it is technically vendor code as it
has been modified rather heavily (as per #8252).

Commit by mmlr on master :: rhrev43636 haiku/: (link)

Add xfreopen.c to the library build, fixes DEBUG build.

The actual xfreopen function is only relevant if O_BINARY != 0 and all
places invoking it are surrounded by if statements that are always
false if O_BINARY == 0. The compiler therefore stripped all the calls
out at the default -O2, but this obviously breaks down with DEBUG on.

Commit by mmlr on master :: rhrev43635 haiku/: (link)

Unconditionally add debug.c. Fixes DEBUG build. [2 commits]

The file is guarded with DEBUG so this doesn't change anything for
non-debug builds. Inspired by patch in #8252 by Urias McCullough.

Commit by mmlr on master :: rhrev43634 haiku/: (link)

Avoid warning on certain debug builds.

The variable is used in the TRACE(), but that is only enabled for
DEBUG >= 2. This therefore avoids a warning on DEBUG=1 builds.

Inspired by patch in #8252 by Urias McCullough.

Commit by mmlr on master :: rhrev43620 haiku/: (link)

Pad the transfer buffer to prevent out of bounds access.

The HIDReportItem reads 32 bit chunks from the report buffer. To avoid
having to check the remaining buffer space on each extraction, we pad
the buffer so it is always valid to read 32 bits at a time. Also add
a comment explaining why we do it that way.

Thanks to Johannes Anderwald for pointing out the potential out of
bound access!

Commit by mmlr on master :: rhrev43510 haiku/: (link)

Fix copy&paste error and double launch of Terminal.

Apparently the "Deskbar" part was a copy&paste error. Also there was
an additional launch call making two Terminal windows come up. Should
fix #8243.

Commit by mmlr on master :: rhrev43484 haiku/: (link)

Fix the build, sorry for the noise.

Commit by mmlr on master :: rhrev43482 haiku/: (link)

Fix use after free in error case. CID 10545.

Commit by mmlr on master :: rhrev43481 haiku/: (link)

Remove extra free() that lead to a double free. CID 10546.

Commit by mmlr on master :: rhrev43480 haiku/: (link)

Remove needless atomic_add() accessing freed memory. CID 10551.

Commit by mmlr on master :: rhrev43478 haiku/: (link)

Fix use after free. CID 10567.

Commit by mmlr on master :: rhrev43476 haiku/: (link)

Fix indirect use of fState after deleting it.

Calling SetIcon(NULL) calls MakeEmpty() that tries to delete all
manipulators on fState. So we may only delete fState after that call.

Also reported as CID 10856.

Commit by mmlr on master :: rhrev43475 haiku/: (link)

Fix build with DEBUG on.

Commit by mmlr on master :: rhrev43473 haiku/: (link)

ASSERT() that there really aren't any wired pages. [2 commits]

Commit by mmlr on master :: rhrev43466 haiku/: (link)

Clean up the mess added in hrev43461 and solve it differently.

• Define a MEMALIGN macro that is either just defined to malloc() or to the actual memalign() depending on where KMessage is used. We only use memalign() inside the kernel and libroot.
• Add a comment to the macro explaining that this allows the use of special heap implementations that might return unaligned buffers for debugging purposes.

Commit by mmlr on master :: rhrev43463 haiku/: (link)

Fix hrev43405 by moving the reference above the autolocker.

The change in hrev43405 wasn't correct, as it put the reference
object definition after the one of the corresponding locker, causing
the reference to be released before the unlock would happen.

Finally fixes #8187. Thanks Ingo for pointing that out.

Commit by mmlr on master :: rhrev43462 haiku/: (link)

Fix invalid use of iterator after erase and lock corruption.

• The call to _TeamDied() causes the team that the iterator points to be removed from the map. Therefore the iterator becomes invalid and may not be accessed anymore (including incrementing it). As we've had to unlock, anything might have happened to to map, so take the safe route and just start over.
• For each dead team that was found the AppManager was unlocked, but there were no balancing lock calls, therefore causing the lock count to get corrupted.