qbit/xenocara
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
3,183 Commits 1 Branch 0 Tags 396 MiB
8c0d9c0f01
Commit Graph

688 Commits

Author SHA1 Message Date
matthieu
9ac8837d36 Make it possible to build drivers with the newer proto headers. ok oga@. 2010-04-13 19:56:37 +00:00
oga
691e04d2ba Here's a really good idea: when preparing for a vt switch the intel 
driver idles the ringbuffer, stops it, unbinds all memory, then hands
back off.

Now here's the clever bit: just before we call to idle the ringbuffer
(which sleeps waiting for an interrupt) let's TURN OFF THE FUCKING
INTERRUPT. That's a great idea, really, isn't it?

Move the interrupt disable to *after* stop_ring. Theo's T61 VT switches
again.  everyone elses' intel stuff worked by magic (or shared
interrupts) before now.

Fixed after discussion with deraadt@, thanks to miod for suggestions.
 2010-04-07 18:22:00 +00:00
matthieu
f53f514d46 Drivers list for loongson. 2010-02-04 06:25:40 +00:00
matthieu
01ee26d198 Recognize WSDISPLAY_TYPE_SMFB 2010-02-04 06:24:47 +00:00
matthieu
5590951830 Remove unused screen_width and screen_height fields from 
driver private data.
 2010-01-10 16:33:44 +00:00
matthieu
73eed35333 Typos in comments and error messages. 2010-01-10 15:22:47 +00:00
matthieu
4fd4d7c316 Default to ZAxisMapping 4 5 6 7. 
This is what most users expect noawadays.
 2009-12-31 18:01:14 +00:00
oga
928f6b4517 Fix a segfault in X for a clipped movie window. 
ok matthieu@

Original commit message from barry scott to upstream:

 Author: Barry Scott <barry.scott@onelan.co.uk>
 Date:   Tue Jun 23 14:14:50 2009 +0100

     Fix segv for clipped movie window

     When playing a movie that is clipped on its left and right edges the
     Xorg server will SEGV sometimes. This is because the intel driver
     ignores the clipping info when it copies the planes out of the XV
     data.  The check for the optimised copy was wrong to ignore the
     width required.  Which leads to too much data being copied by the
     memcpy. It the source buffe happens to end exactly on a page
     boundary the server will SEGV.

     As we reviewed the code we checked the calculation of src1, src2 and
     src3.  The patch includes additional comments to make it clear what
     the elements of the calculation are.

     This bug exists in git head and we also see it in 2.4.1.
 2009-12-14 14:21:54 +00:00
matthieu
1f09a7b580 Link xf86-video-wsudl to the build on architectures 
that support udl(4). ok todd@
 2009-12-12 09:46:40 +00:00
matthieu
012236a1be Fix SwapXY and rotation support. 
More work to do wrt properties and calibration.
 2009-12-10 22:32:02 +00:00
oga
80bba657d4 split the KB_VARTAB code in wscons autodetection so that it understands 
both XkbOptions and XkbVariant. now we can understand the .dvorak,
.nodeadkeys and .swapctrlcaps wscons variants and pass them correctly onto XKB.
Means that xmodmap or setxkbmap magic isn't needed in those cases
anymore.

"no objections" matthieu@. Diff from Tobias Ulmer (tobiasu at tmux dot
org); thanks!
 2009-12-07 19:28:53 +00:00
matthieu
bf862c28bf Remove the input driver before returning NULL on error in PreInit(). 2009-12-06 10:32:31 +00:00
matthieu
dfdb9035a8 update to xf86-video-mga 1.4.11. Tested by landry@ 2009-11-29 20:10:04 +00:00
matthieu
c98180f2b7 missing commit for mid-button emulation. 2009-11-27 10:30:08 +00:00
matthieu
5f158e3016 Cleanup this driver a bit, following what's been done in xf86-input-ws: 
- no more convert proc
- use the correct XInput device type
- don't hardcode the package version number in the .c file.
 2009-11-26 21:30:19 +00:00
matthieu
dcbd99cf98 white space 2009-11-26 18:13:47 +00:00
matthieu
bdbb8460ce regen 2009-11-26 16:45:41 +00:00
matthieu
007942360f Missed this file when adding middle button emulation. 2009-11-26 16:45:02 +00:00
matthieu
8b15b85686 Documentation for Emulate3Buttons and Emulate3Timeout and the associated 
properties.
 2009-11-26 16:42:52 +00:00
matthieu
c9960a2b5f Add middle mouse button emulation to xf86-input-ws. 
Code stolen from evdev, that stole it from xf86-input mouse.
 2009-11-26 16:42:06 +00:00
matthieu
18f9c26208 Document device properties 2009-11-26 11:10:09 +00:00
matthieu
910b2d77db Set resx/resy fields of the wscons calibration structure to the saved 
values when the calibration properties change.
 2009-11-26 10:57:35 +00:00
matthieu
cfe9a6b607 Only default to raw mode for touchpanels. The raw mode doesn't make 
sense for mices.
 2009-11-25 19:36:57 +00:00
matthieu
e945969695 When calibrating directly update the axis structures with new min/max values. 
In xserver 1.7 with axis labels, the init function used previously
makes no sense anymore.
 2009-11-25 19:31:35 +00:00
matthieu
3147c5ee18 The raw mode is a lot more useful now. So enable it by default. 2009-11-25 19:11:55 +00:00
matthieu
d5dc0ce8f5 No need to maintain a local cache of the pointer position anymore. 2009-11-25 18:14:23 +00:00
matthieu
d28a98a22a regen 2009-11-25 18:12:02 +00:00
matthieu
6665646a28 Implement device properties for calibration data. 
Setting the properties also write to the kernel table.
The calibration code can now change the data without quitting X.
 2009-11-25 18:10:26 +00:00
matthieu
15298ef907 Save a copy of the kernel calibration data in the private data and 
use it to save/restore the state during DEVICE_ON/DEVICE_OFF.
 2009-11-25 18:03:42 +00:00
matthieu
fcd31496fc Send events upstreams one axis at a time. 2009-11-25 17:59:42 +00:00
matthieu
279e764019 Update to xf86-video-radeonhd 1.3.0 
Tested by killi@ on a HD3200.
 2009-11-24 16:57:34 +00:00
matthieu
c3d957f5fb Update to xf86-video-openchrome 0.2.904. 
Tested by myself and kevlo@ on V800 and V855 chipsets.
 2009-11-24 16:54:55 +00:00
matthieu
2baec57bee regen 2009-11-23 22:53:39 +00:00
matthieu
0e46d3a279 Missed one openbsd->xf86 rename. 2009-11-23 22:52:38 +00:00
matthieu
a04d9756c9 conversion_proc is not used anymore since Xserver 1.5. 
Don't define one that's never called.
 2009-11-23 22:10:29 +00:00
matthieu
bd0a1d71f5 Document option "raw". 2009-11-23 21:33:11 +00:00
matthieu
91bbfc2c74 Fix previous for non tablet cases. the WSMOUSE_IO_GCALIBCOODS fails there. 2009-11-23 21:11:27 +00:00
matthieu
76bf8ac9b1 Implement a new "raw" mode that will put a panel in raw mode. 
Calibration data is read from the kernel, so things still work in this
mode.
 2009-11-23 18:29:13 +00:00
matthieu
c56921a15d Another unused declaratin bites the dust. 2009-11-23 17:36:23 +00:00
matthieu
068d81918d - fix parameter orders for XInput 2 xf86InitValuatorAxisStruct() calls 
- set min/max axis values for absolute valuators
 2009-11-23 16:55:52 +00:00
matthieu
feea0e6d5e Set buttons and axis labels for XInput 2. 2009-11-23 16:21:50 +00:00
matthieu
fd1ce4e8f3 Handle NetBSD's wsmouse ABI versions. From marc at msys dot ch. Thanks. 2009-11-23 15:25:11 +00:00
matthieu
96a4b61607 split wsProc(). 2009-11-23 15:16:52 +00:00
matthieu
8dd19e6aa8 typo. From marc@msys.ch. Thanks. 2009-11-23 14:00:17 +00:00
matthieu
d0d883f971 regen 2009-11-23 12:55:49 +00:00
matthieu
39f6c2f761 Bump version -> 1.2.9 2009-11-23 12:54:38 +00:00
matthieu
1d2af3abc7 Fetch the module version from autotools instead of hard-coding it. 2009-11-23 12:54:04 +00:00
matthieu
f8b25a1513 kill white space at EOL 2009-11-23 12:37:32 +00:00
matthieu
e9b14fe7a9 Set correct device type and mode in XInput structures. 2009-11-23 12:35:40 +00:00
matthieu
3a858b5a6e unifdef USB_NEW_HID. 
USB_GET_REPORT_ID is available on NetBSD and OpenBSD for more
than 7 years now.
 2009-11-22 23:16:54 +00:00
matthieu
d1e2bf33ac Get rid of the now unused remaining moduleInfoRec bits. 2009-11-22 23:06:24 +00:00
matthieu
1403cc1e0c regen 2009-11-22 23:04:47 +00:00
matthieu
bbabd84556 Update to idioms used in modern X.Org input drivers. 2009-11-22 23:03:04 +00:00
matthieu
53fe2d0ea0 Get rid of the now unused remaining moduleInfoRec bits. 2009-11-22 23:00:43 +00:00
matthieu
876c2ea266 Prepare for ABI_XINPUT_VERSION 7. 
Still missing the actual button and axis labeling,
this can be added at a later point in time.
 2009-11-22 22:30:18 +00:00
matthieu
3f1a5f5c59 unifdef XFree86Loader. It's mandatory nowadays. 2009-11-22 22:20:26 +00:00
matthieu
68823f5b52 Update to idioms used in modern X.Org input drivers. 2009-11-22 22:18:12 +00:00
matthieu
deae15d61d Prepare for ABI_XINPUT_VERSION 7. 
Still missing the actual button and axis labeling,
this can be added at a later point in time.
 2009-11-22 22:08:57 +00:00
matthieu
d8fba39f32 regen 2009-11-22 19:40:18 +00:00
matthieu
c6e9475115 Update to idioms used in modern X.Org input drivers. 2009-11-22 19:39:03 +00:00
matthieu
799a140adf unifdef XFree86Loader. It's mandatory nowadays. 2009-11-22 19:26:48 +00:00
matthieu
bd717db836 Renamed two driver which had incorrect names. 
(The 1st part of an Xorg driver name is the name of the DDX,
not the one of your preferred operating system).
 2009-11-22 19:16:48 +00:00
matthieu
1e2202c25f Re-import the wscons pointer input driver using the correct name. 2009-11-22 19:12:44 +00:00
matthieu
84fbe7508a Re-import the usbtablet driver using the correct name. 2009-11-22 19:10:35 +00:00
matthieu
098607ee0a Also update ChangeLog from upstreams. 2009-11-22 18:51:24 +00:00
matthieu
7a884b74e6 Update for resources/RAC API removal and bump to 0.3.0 to follow 
upstreams numbering.
 2009-11-22 18:48:19 +00:00
matthieu
8af4b716e8 Update to xf86-video-voodoo 1.2.3 2009-11-22 18:29:21 +00:00
matthieu
8b6a5661b0 Update to xf86-video-vmware 10.16.8 2009-11-22 18:25:28 +00:00
matthieu
56ccae54cf Update to xf86-video-vesa 2.2.1 2009-11-22 18:21:12 +00:00
matthieu
bdaf8eccf3 Update to xf86-video-tseng 1.2.3 2009-11-22 18:18:06 +00:00
matthieu
1269050414 Update to xf86-video-trident 1.3.3 2009-11-22 18:13:13 +00:00
matthieu
a225cbaffd Update to xf86-video-tga 1.2.1 2009-11-22 18:09:24 +00:00
matthieu
9134425e67 Update to xf86-video-tdfx 1.4.3 2009-11-22 18:05:12 +00:00
matthieu
9fec9b5dce Remove the sunbw2 driver which has been deprecated upstreams and 
was never built on OpenBSD.
 2009-11-22 17:58:22 +00:00
matthieu
7ee5b3111c Update to xf86-video-sisusb 0.9.3 2009-11-22 17:56:41 +00:00
matthieu
a8596e56ea Update to xf86-video-sis 0.10.2 2009-11-22 17:54:56 +00:00
matthieu
61ff9ac41f Update to xf86-video-siliconmotion 1.7.3 2009-11-22 17:53:03 +00:00
matthieu
4122cf8c50 Update to xf86-video-savage 2.3.1 2009-11-22 17:48:10 +00:00
matthieu
180745e43a Update to xf86-video-s3virge 1.10.4 2009-11-22 17:43:07 +00:00
matthieu
9f1b635ef5 Update to xf86-video-s3 0.6.3 2009-11-22 17:37:38 +00:00
matthieu
4120b53167 Update to xf86-video-rendition 4.2.3 2009-11-22 17:31:53 +00:00
matthieu
a0c49ba554 Update to xf86-video-r128 6.8.1 2009-11-22 17:11:19 +00:00
matthieu
ddc8101056 Update to xf86-newport 0.2.3 2009-11-22 16:54:09 +00:00
matthieu
453dcc622b Update to xf86-video-neomagic 1.2.4 2009-11-22 16:50:00 +00:00
matthieu
c526cc488e Update to xf86-video-i740 1.3.2 2009-11-22 16:10:56 +00:00
matthieu
13cbc50e1d Update to xf86-video-dummy 0.3.2 2009-11-22 16:05:45 +00:00
matthieu
aec1eae93c Update to xf86-video-cirrus 1.3.2 2009-11-22 16:01:56 +00:00
matthieu
eccc1caedd Update to xf86-video-chips 1.2.2 2009-11-22 15:55:20 +00:00
matthieu
b70d13b105 Update to xf86-video-ast 0.89.9 2009-11-22 15:33:58 +00:00
matthieu
c9291cc9ce Update xf86-video-ark to 0.7.2 2009-11-22 15:29:04 +00:00
matthieu
2c476df1f0 Update xf86-video-apm to 1.2.2 2009-11-22 15:24:48 +00:00
matthieu
10a4e47133 Update to xf86-input-joystick 1.5.0 2009-11-22 14:38:31 +00:00
matthieu
c25d6a0103 Update xf86-input-void to version 1.3.0 2009-11-22 14:32:27 +00:00
matthieu
a391d7d7b4 Update xf86-input-vmmouse to version 12.6.5 2009-11-22 14:29:51 +00:00
matthieu
04895411e3 update xf86-input-acecad to 1.4.0. 2009-11-22 14:25:33 +00:00
matthieu
a97527f412 pass the real framebuffer depth to xf86SetDepthBpp() instead of letting 
it guess. This lets xf86-video-wsfb on TGA framebuffers on alpha
whithout setting weights and fbpp explicitely.
 2009-10-19 20:01:30 +00:00
matthieu
47491b577c update to xf86-video-geode 2.11.6. Tested on a Alix 3C3. 2009-10-11 09:30:14 +00:00
matthieu
c31b066623 update to xf86-video-nv 2.1.15. Tested by espie@ (not better, 
but not worse either on his hardware).
 2009-10-10 18:17:43 +00:00
matthieu
eff17519e4 Update for newer Xinput ABI. 2009-10-03 10:17:04 +00:00
mglocker
57c0324fbe Adapt to the new UDLIO_DAMAGE ioctl behaviour of udl(4). 2009-09-27 18:19:28 +00:00
mglocker
f79025b4ae wsudl(4) / udl(4) works on NetBSD as well now. 
From FUKAUMI Naoki
 2009-09-24 07:34:40 +00:00
mglocker
21fe69d289 If udl is to busy to process our UDLIO_DAMAGE drawing request, let wsudl 
sleep for a bit before retrying, so we don't overload the kernel by busy
looping.

Tested and agreed matthieu@
 2009-09-23 20:09:03 +00:00
matthieu
aa2fb42def XF86Config -> xorg.conf in comments. 2009-09-22 22:00:54 +00:00
mglocker
1360291edb Initial Xorg driver to support DisplayLink USB display devices using 
the udl(4) device driver over wsdisplay(4).

OK matthieu@, miod@
 2009-09-22 14:11:43 +00:00
matthieu
e10bf73c09 Same fix as in wsfb: 
Call shadowRemove() in wsfbCloseScreen(). This fixes a free
memory dereference on exit. Hint by Daniel Stone. Thanks.
 2009-09-13 19:46:47 +00:00
matthieu
5331f4bab2 Call shadowRemove() in wsfbCloseScreen(). This fixes a free 
memory dereference on exit. Hint by Daniel Stone. Thanks.
 2009-09-13 19:33:49 +00:00
oga
f0cd9e3a73 backport a bugfix from upstream: 
Fix 945GM VT switch in UMS

	Bug #19578. We should set private intel_crtc state according to
	current, as fail to do so pipe A needs active won't be taken
	care of. Also make sure pipe swap operation always set during VT
	switch.

This also fixes the *other* x40 lid problem (which happens when you suspend,
then you shut the lid in dont'-suspend-on-lid mode.

the reasoning for the PipeA quirk I finally found:

	On some platforms, the firmware may read & write GPU registers
	on lid close, suspend/resume time or during various SMM events.
	If one of the graphics pipes is disabled at that time, the GPU
	may hang due to the programming dependencies of the various
	registers.

So essentially, without this patch, if you have this quirk and you shut
the lid, you're hosed, but y'know drivers in userland are great. So are
BIOS writers.

requested by deraadt@, who keeps getting this issue when he's on a plane.
 2009-09-10 02:30:26 +00:00
matthieu
b1c6b32ab5 On powerpc, only set "UseFBDev" option to true by default on Linux. 2009-09-06 15:45:43 +00:00
matthieu
f090af278c MFC: enable build with recent input ABI. 2009-09-06 12:22:42 +00:00
matthieu
7370432720 Remove compat directories that are not needed here. ok oga@. 2009-08-25 19:08:29 +00:00
matthieu
f74c2dba55 update do xf86-video-ati 6.12.2 
This has been in snapshots for weeks. ok oga@, todd@.
 2009-08-25 18:51:44 +00:00
oga
5f44320c03 Pull in some code from 2.8 and current so that dri2 works correctly with 
xserver 1.6.2. When dri2 is not in use, this commit does nothing.

ok matthieu@
 2009-08-06 16:02:07 +00:00
kettenis
276db15ee6 Disable hardware cursor before we exit. From NetBSD. 
ok oga@
 2009-07-19 14:46:59 +00:00
matthieu
07600073fd Correct patch for libpciaccess issue from X.Org bug 18122, by 
Marien Zwart. Tested by Pierre Cardoso.
 2009-07-18 20:16:00 +00:00
oga
9a5da84da5 Since we don't accelerate trapesoid rendering under uxa yet anyway 
software rasterise them THEN put them into buffer objects.
read-modify-write of graphics memory is expensive.

2.5x speed increase under heavy trap rendering measured.

Another upstream backport. ok matthieu@
 2009-07-18 15:02:51 +00:00
oga
2b56f07bde Don't put glyphs pixmaps into buffer objects under uxa. 
in the fastpath they're going straight to the glyph cache anyway and
doing so for small glyphs wastes memory since bos must be padded to page
size.

Backport from upstream (to aid gem development). ok matthieu@
 2009-07-18 15:00:09 +00:00
oga
c349058b6d Fallback if we're vt switched and we're using GEM. We're not allowed to 
access the hardware in this situation.

Backport from upstream CURRENT.

ok matthieu@.
 2009-07-18 14:58:00 +00:00
oga
37a2f39d00 Accessing a pixmap in fallback if the prepare_access() callback failed 
is verboten

So don't do it. Treat it the same as every other prepare access call in
uxa.  This patch sent upstream.

ok matthieu@
 2009-07-18 14:56:03 +00:00
matthieu
4d65740941 Fix PCI card probe and chipset detection. 
it appears that the chips&technologies X driver was partially broken
during libpciaccess conversion. Thanks to the analysis by Marien Swart
in bug #18122 and a similar fix by Mark Kettenis for the radeon
driver, this patch should fix it.
 2009-06-30 18:39:19 +00:00
matthieu
ffd83bca19 update to svn revision 753. This includes a few bug fixes that 
make EXA work on my Chrome9 HC3 IGP. ok oga@
 2009-06-27 10:11:47 +00:00
matthieu
e70d43c060 I was supposed to remove the --enable-debug option before committing. 2009-06-25 21:28:57 +00:00
matthieu
855916fc83 update to openchrome.net svn commit 750. 
This adds among others support for the VX800 chipset.
This has been in snapshots for weeks.
 2009-06-25 21:27:18 +00:00
matthieu
42b052b942 update to xf86-video-intel 2.7.1. Tested by many. 2009-06-25 20:16:43 +00:00
matthieu
810f9ab0c6 Don't de-reference a NULL pointer in R128FreeScreen in the error path. 
If R128PreInit() failed, R128FreeRec() was already called and info
will be NULL.
 2009-06-13 17:18:21 +00:00
matthieu
787b897b8c Remove the xf86-video-suncg6 driver, which doesn't support acceleration 
beside the hw cursor. The latter can be done in wsfb if someone manages
to prove miod@ that it's useful.
 2009-06-12 19:59:00 +00:00
matthieu
dd6a77af66 Remove the input drivers that have been dropped by X.Org and 
were not built for months now.
 2009-06-05 21:26:36 +00:00
matthieu
2b279ed91b Update to xf86-video-sis 0.10.1. 2009-06-05 21:05:04 +00:00
matthieu
34be51c4e7 Update to xf86-video-suntcx 1.1.1 2009-06-05 20:50:02 +00:00
matthieu
ae73f8111a regen 2009-06-05 20:47:48 +00:00
matthieu
c6aa2d45af update to xf86-video-suncg6 1.1.1 2009-06-05 20:46:32 +00:00
matthieu
4bb62c07fc update to xf86-video-suncg3 1.1.1 2009-06-05 20:44:42 +00:00
matthieu
a58e80c5a2 regen 2009-06-05 20:42:19 +00:00
matthieu
d5f37b904b update xf86-video-newport to 0.2.2 2009-06-05 20:40:49 +00:00
kettenis
3bb964735c Don't set the default depth to 8bpp. These days 24/32bpp makes much more 
sense.

ok matthieu@
 2009-06-04 22:54:25 +00:00
kettenis
8e04fb5a97 Disable int10 code if __sparc__. 
ok matthieu@
 2009-06-04 21:04:57 +00:00
kettenis
ecf487b85b Build xf86-video-glint on OpenBSD/sparc64. 2009-06-04 18:36:30 +00:00
kettenis
3ea6bb7f0e Map and unmap registers around DDC probe. 
ok matthieu@
 2009-06-04 18:20:10 +00:00
kettenis
86fd76579f The mmio registers consist of two blocks of 64k. The first 64k provide 
a little-endian view, the second 64k provide a big-endian view.  Restrict
mapping to 64k to prevent mapping beyond the end of the BAR if we map the
big-endian view on big-endian machines.  There is no reason to map the
big-endian view on little-endian machines.

ok matthieu@
 2009-06-04 18:18:11 +00:00
kettenis
037d9f8966 regen 2009-06-04 17:28:08 +00:00
kettenis
de58239bc4 Since the driver uses the big-endian view of the registers we shouldn't be 
byteswapping during register access.  Define PPC_MMIO_IS_BE and
SPARC_MMIO_IS_BE to prevent this.

ok matthieu@
 2009-06-04 17:21:00 +00:00
matthieu
4579839317 regen 2009-06-01 21:30:04 +00:00
matthieu
0cce1f2f9a Update to be compatible with xserver 1.6 and Xinput 1.5. 2009-06-01 21:29:06 +00:00
matthieu
70a4c30fcb regen 2009-06-01 21:18:08 +00:00
matthieu
05f27001f1 Update to be compatible with xserver 1.6 and Xinput 1.5. 2009-06-01 21:15:50 +00:00
matthieu
1dd9fa7be4 regen 2009-06-01 20:58:16 +00:00
matthieu
dc5e2c7159 Don't call xf1bpp or xf4bpp if they're not present (xorg-server-1.6 & later) 
From  Alan Coopersmith <alan.coopersmith@sun.com>.
 2009-06-01 20:57:20 +00:00
matthieu
5e0c79c5ff update to xf86-video-vmware 10.16.6 2009-05-18 20:58:46 +00:00
matthieu
29dfc1a9fe update to xf86-input-vmmouse 12.6.4 2009-05-18 20:53:43 +00:00
kettenis
c323e0e8a5 Set sane defaults fori the reference clock for Rage XL and Mobility cards 
on sparc64.  This makes it possible to run X without an xorg.conf on machines
like the blade100 and the Tadpole laptops.

ok oga@, matthieu@
 2009-05-04 20:34:28 +00:00
kettenis
d69ace73ff Fix a problem introduced by the libpciaccess conversion where the chip ID 
would be set wrong and the proper resources wouldn't be claimed.

ok oga@, matthieu@
 2009-05-04 20:32:21 +00:00