Ros-diffs August 2009

ros-diffs@reactos.org
  • 21 participants
  • 654 discussions

[cgutman] 42384: - Add a debug print
by cgutman@svn.reactos.org
15 years, 1 month

[gschneider] 42383: _ecvt: - Rewrite the function, getting rid of alloca and unneeded code parts - Relies on working sprintf, especially allows to detect the decimal point position independent of wanted precision - Fixes all msvcrt printf winetests
by gschneider@svn.reactos.org
15 years, 1 month

[gschneider] 42382: sprintf: - Implement higher precision exponent selection, checked against ecvt test cases to match wanted behavior - msvcrt printf winetests up by ~10 because current ecvt relies on non-correct behavior - Cleanup unneeded checks, wrong variable usages, add new header
by gschneider@svn.reactos.org
15 years, 1 month

[fireball] 42381: - Uncomment missing surface unlocks.
by fireball@svn.reactos.org
15 years, 1 month

[fireball] 42380: - Use a precreated stock bitmap.
by fireball@svn.reactos.org
15 years, 1 month

[fireball] 42379: - Connect a new Win32 process to the window station, as it's supposed to be done. However, desktop inheritance still needs to be checked. There is some testing code in connect_process_winstation left #if0ed out.
by fireball@svn.reactos.org
15 years, 1 month

[sginsberg] 42378: - Remove asm-optimized versions of RtlRandom and RtlUniform -- there is *no* reason to use hand-optimized assembly for simple routines like these -- let the compiler do its job, please. Also, we compiled random.c *and* random_asm.S, meaning we duplicated the functions -- the compiler didn't detect this and silently used the C versions. - Don't forward RtlRandomEx to RtlRandom in ntdll -- RtlRandomEx should use a different algorithm (but is simply a copy of RtlRandom right now)
by sginsberg@svn.reactos.org
15 years, 1 month

[fireball] 42377: - Merge 42376 (along with committing a merge history leftover from a previous merge).
by fireball@svn.reactos.org
15 years, 1 month

[fireball] 42376: - Add PsGetProcessInheritedFromUniqueProcessId declaration to NDK.
by fireball@svn.reactos.org
15 years, 1 month

[cgutman] 42375: Broadcast fixes - Treat 255.255.255.255 as an undefined address - Choose the first interface when sending a broadcast packet (fixes failing with STATUS_NETWORK_UNREACHABLE when trying to send a broadcast packet) - Fix a broadcast address check so 255.255.255.255 will also pass (fixes sending queued broadcast packets) - Now a broadcast packet can be successfully sent from an interface with a valid IP address (previously they could only be sent when the interface didn't have an NC
by cgutman@svn.reactos.org
15 years, 1 month
Results per page: