summaryrefslogtreecommitdiffstats
path: root/sys/dev/usb/udl.c (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* If UDLIO_DAMAGE fails to queue the damage drawing request, undo themglocker2009-09-191-7/+14
| | | | operation and tell the X11 driver to retry later (same as for rasops).
* Add bits for X11 support over the wsdisplay framework (fully Huffmanmglocker2009-09-191-2/+143
| | | | | | | compressed). The upcoming X11 driver mmaps its framebuffer from the kernel and sends updates via the Damage mechanism. Committed from a X session on udl, of course.
* Remove (last) obsolete panic() case.mglocker2009-09-131-3/+3
|
* Fix command undo for compression. Fixes artifacts seen in some cases.mglocker2009-09-131-25/+21
|
* Move some real noisy operations to a higher debug level.mglocker2009-09-131-4/+5
|
* Use function pointers in erasecols and eraserows for framebuffermglocker2009-09-131-3/+3
| | | | operations, so we do compression in there as well.
* Don't use function pointers at places where we always call the samemglocker2009-09-121-2/+2
| | | | function.
* Re-sort some functions. Fix comment. No functional changes.mglocker2009-09-121-197/+198
|
* Don't panic if the command buffer runs full. Instead try to flush it,mglocker2009-09-111-84/+225
| | | | | | | | | and continue operation afterwards. If we can't flush immediately, we will undo the operation and return EAGAIN to wsdisplay (like we do already for the xfer queue). This allows us to set higher resolutions (tested up to 1280x1024) without bumping the command buffer or xfer buffer memory further.
* Today we just support a font width up to 8 pixels. Allow a font widthmglocker2009-09-091-5/+10
| | | | | | up to 16 pixels now. Reported by maja@
* Fix typo in comment, safe -> save.mglocker2009-09-061-3/+3
| | | | Spotted by miod@
* Fix udl_copycols overlay copy problem by copying area to off-screen firstmglocker2009-09-061-2/+6
| | | | (same as in udl_copyrows).
* Make use of the new wsdisplay ability, committed by miod previously, whichmglocker2009-09-051-25/+77
| | | | | | | | | allows our rasops functions to return EAGAIN if our usb command queue is full. This gets us rid of the ugly while { delay(); } loop and makes udl work also on MP kernels now (we faced a deadlock so far because spinning in that delay() loop with biglock active wasn't that much fun). With help from miod@
* Change the wsdisplay_emulops return types from void to int; emulops will nowmiod2009-09-051-13/+25
| | | | | | return zero on success and nonzero on failure. This commit only performs mechanical changes for the existing emulops to always return zero.
* Add some new devices. -mojmaja2009-08-301-2/+4
|
* Ooops, revert the last bcopy()-elimination commit. I forgot that bcopy()mglocker2009-08-291-7/+7
| | | | does the job for us on strict alignment archs.
* Reduce padding overhead in compressed blocks.mglocker2009-08-271-10/+16
|
* Remove obsolete comment.mglocker2009-08-261-3/+2
|
* Replace some stupidly used bcopy()s by direct assignment.mglocker2009-08-261-8/+8
|
* Also avoid __packed struct to Huffman table in the driver side in favourmglocker2009-08-261-3/+3
| | | | | | of performance. Instead generate the Huffman table with two ints. Suggested, help and OK deraadt@
* Enable pixel compression (Huffman pixel difference method). If themglocker2009-08-251-10/+419
| | | | | Huffman table can't be loaded, the driver will operate in uncompressed mode.
* Check if the command buffer has been allocated before we free it. Fixesmglocker2009-08-091-2/+5
| | | | a possible page fault trap while device detach.
* Never return nonzero in a device activate method invoked with DVACT_ACTIVATE,miod2009-08-021-2/+2
| | | | | for this prevents it to be invoked with DVACT_DEACTIVATE later. This had been sweeped some time ago already, but bad constructs crept in again.
* Add some more udl devices. -mojmaja2009-07-231-2/+6
|
* add another UDL deviceyuo2009-06-061-1/+2
| | | | | | - LCD-4300U (century corp.) ok mglocker@
* Use the background color attribute in udl_erasecols and udl_eraserowsmglocker2009-06-011-3/+10
| | | | instead hardcoding ``0x0000''.
* Sort argument list of udl_draw_char similar as udl_fb_* friends.mglocker2009-06-011-6/+6
|
* Allow a font width other than 8 pixels.mglocker2009-06-011-2/+2
|
* Better naming for some framebuffer functions.mglocker2009-06-011-21/+21
|
* Reduce data payload for font drawing to a fixed size per character.mglocker2009-06-011-9/+31
|
* I forgot to apply the End Of Command token for async xfers. Maybe thatmglocker2009-06-011-1/+5
| | | | caused the stalls seen in some rare conditions? Who knows ...
* We are able to address 256 pixels (16bit) per command not just 128 pixels.mglocker2009-05-311-8/+8
| | | | | This safes us some protocol overhead, especially when bit blittering large regions.
* Mention broken cursor drawing.mglocker2009-05-241-1/+9
|
* List product name instead chip type.mglocker2009-05-241-3/+3
|
* Add Unitek Y-2240. -mojmaja2009-05-221-1/+2
|
* Add König CMP-USBVGA10. -mojmaja2009-05-211-1/+2
|
* Add ForwardVideo EasyCAP008. -mojmaja2009-05-211-2/+3
|
* Add support for the DisplayLink USB<->DVI Adapter.mglocker2009-05-181-2/+3
|
* In debug mode let us know which ioctl has been requested.mglocker2009-05-101-2/+3
|
* Make inactivity screen blanking (burner) work.mglocker2009-05-101-6/+6
|
* Adjust debug level to be not that noisy by default (level 1).mglocker2009-05-101-4/+4
|
* Return -1 from unused udl_mmap() function instead of 0.mglocker2009-05-091-2/+2
| | | | Spotted by miod@
* Initial console device driver for USB displays based on the DisplayLinkmglocker2009-05-091-0/+1344
DL-120 / DL-160 graphic chips, using the wsdisplay(4) layer. Based on the reversed engineered specifications of Florian Echtler. The driver is disabled in GENERIC for now. Thanks to claudio@ for donating me a device and to oga@ and miod@ for giving me hints in the graphic and wsdisplay area. OK deraadt
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.