Don't crash when updating menu item that's not in the window server. Menu Items begin with an invalid ID of -1, and only get a valid ID if the menu they are attached to is added to the window server. They don't send updates to the server unless they have a valid ID.
Fixes#82
fixed it so that directories don't get double-opened (first when they are
opened, and second when the selection changes to match in the file tree view)
This is in preparation for eventually using it in userspace.
LinearAddress.h has not been moved for the time being (as it seems to be
only used by a very small part of the code).
Since we transition to a new PageDirectory on exec(), we need a matching
RangeAllocator to go with the new directory. Instead of juggling this in
Process and MemoryManager, simply attach the RangeAllocator to the
PageDirectory instead.
Fixes#61.
This gives us some leeway for WindowServer to queue up a bunch of messages
for one of its clients. Longer-term we should improve DoubleBuffer to be
able to grow dynamically in a way that gets billed to some reasonable place.
Passing this flag to recv() temporarily puts the file descriptor into
non-blocking mode.
Also implement LocalSocket::recv() as a simple forwarding to read().
Oops, it looks like I left the max inline rects limit at 1 while debugging
this code.. We can fit 32 rects in a single WSAPI message without needing
a second "extra data" message, so let's use the space we have!
Move the bulk of exec() into a new pump(). Since SDL wants to drive the
event loop itself, this is a requirement. We also add a WaitMode flag to
allow for immediately pumping events -- again, this is required because
SDL wants to be in full control of the event loop, and not let us wait.
can_write() was saying yes in situations where write() would overflow the
internal buffer. This patch adds a has_attached_peer() helper to make it
easier to understand what's going on in these functions.
This is not EOF, and never should have been so -- can trip up other code
when porting.
Also updates LibGUI and WindowServer which both relied on the old
behaviour (and didn't work without changes). There may be others, but I
didn't run into them with a quick inspection.
Rather than having the first click hit a bomb, if the first click would
hit a bomb, instead, reset the game board.
This is a (sort of) feature of Windows minesweeper, and IMO makes
playing a bit more fun :-)
* EPIPE now correctly deletes the client connection
* EAGAIN (which is now returned by the kernel if the write buffer fills)
terminates the connection also
