4e91ae275e
Simplify: Always use writev. (In _xcb_out_flush, convert the output queue to a single iovec if needed.)
2006-02-24 00:02:43 -08:00
7f0bc778c8
Factor padding out of _xcb_out_write_block and into its callers, XCBSendRequest and write_setup.
...
This requires dynamically allocating memory in XCBSendRequest, but this
malloc/free pair turns out to cause a 30% speed hit for the 'x11perf -noop'
test -- so for the moment I use alloca where available and fall back to malloc
on other platforms. Later I think I'll change the contract of XCBSendRequest
so the caller is responsible for memory allocation, because the caller ought
to always be able to stack-allocate here.
2006-02-23 22:41:59 -08:00
6e29e5f2ee
Add XCBGetQueuedRequestRead for Xlib that does no syscalls, just returns whatever XCB already knows about.
2006-02-23 17:50:53 -08:00
26ac6292ba
Minor performance fix: Change the contract on XCBGetRequestSent so that it does not waste time re-locking. This is for Xlib, and Xlib has already locked.
2006-02-23 15:29:40 -08:00
a736674943
Minor performance fix: Only rearrange buffers for BIG-REQUESTs.
2006-02-23 14:57:46 -08:00
e866bed934
Move request_written update back where it was for now: doing it early can cause XCBWaitForReply to wrongly believe that the request has been flushed. Eventually, we should fix bug #6021 .
2006-02-23 14:32:11 -08:00
1b83f8f8f3
Remove _xcb_assert_valid_sequence. One test is trivially true, and the other may be temporarily violated without anything bad happening.
2006-02-23 14:30:08 -08:00
55c1842686
Move _xcb_write and _xcb_writev to xcb_out.c and make them static, since only _xcb_out_write calls them.
2006-02-23 12:48:27 -08:00
213b572592
Ignore test byproducts.
2006-02-23 12:34:08 -08:00
6149c7a6b5
More return value changes, and make _xcb_in_read_packet static since it is not called from outside xcb_in.c.
2006-02-23 12:15:09 -08:00
5b1d39e27b
More return value changes.
2006-02-23 11:50:12 -08:00
662479760c
Make the return value of _xcb_conn_wait boolean, instead of syscall-like.
2006-02-23 10:17:40 -08:00
f8a8b465e0
Bugfix: move request_written update *before* _xcb_conn_wait in _xcb_out_flush. Otherwise a reply may be read before we record that we have sent the request, and then XCB gets... confused.
2006-02-22 23:38:16 -08:00
e0a3578383
Remove USE_THREAD_ASSERT compile option. _xcb_conn_wait can no longer be re-entered, since we have no callback hooks now.
2006-02-22 23:11:36 -08:00
a7d749ec3f
Remove vestiges of adjacent request combining implementation. It is very dead.
2006-02-22 23:09:29 -08:00
06c788932b
Control debugging, optimization, and warning flags all independently.
2006-02-20 00:41:39 -08:00
3f98dae4a8
Merge branch 'master' of git+ssh://git.freedesktop.org/git/xcb
2006-02-18 23:57:18 -08:00
e346f8c8ef
Update .gitignores for .o files and autofoo stuff.
2006-02-18 23:57:14 -08:00
f825a32135
Add missing entries to .gitignore files.
2006-02-18 23:05:32 -08:00
9e013b564c
Merge branch 'master' of git+ssh://anholt@git.freedesktop.org/git/xcb
2006-02-18 22:57:42 -08:00
d16efb5191
Move .cvsignore to .gitignore.
2006-02-18 22:56:03 -08:00
7474473f33
Quit defining _XOPEN_SOURCE. We never needed it...
2006-02-18 20:03:35 -08:00
1d977c7b2d
Remove the old ChangeLog file. GIT makes that obsolete.
2006-02-18 19:23:33 -08:00
5b7182c659
Workaround X server bug, fd.o #3210 : if a GLXGetFBConfigs request is sent, fix the length field in the reply.
2006-02-18 18:12:40 -08:00
47ceed7861
Remove xcl and CVSROOT.
2006-02-18 16:49:41 -08:00
Jamey Sharp
Eric Anholt
Jamey Sharp
Josh Triplett