* Re: [GIT PULL FOR 2.6.37] new AF9015 devices
2010-09-09 21:12 [GIT PULL FOR 2.6.37] new AF9015 devices Antti Palosaari
@ 2010-09-09 23:19 ` Stefan Lippers-Hollmann
2010-09-10 0:54 ` Stefan Lippers-Hollmann
2010-09-22 19:02 ` Mauro Carvalho Chehab
2 siblings, 0 replies; 6+ messages in thread
From: Stefan Lippers-Hollmann @ 2010-09-09 23:19 UTC (permalink / raw)
To: Antti Palosaari; +Cc: Mauro Carvalho Chehab, linux-media, TerraTux
Hi
On Friday 10 September 2010, Antti Palosaari wrote:
> Moikka Mauro!
> This patch series adds support for TerraTec Cinergy T Stick Dual RC and
> TerraTec Cinergy T Stick RC. Also MxL5007T devices with ref. design IDs
> should be working. Cinergy T Stick remote is most likely not working
> since it seems to use different remote as Cinergy T Dual... Stefan could
> you test and ensure T Stick is working?
>
> and thanks to TerraTec!
[...]
DVB-T is working fine with my TerraTec "Cinergy T RC MKII" 0ccd:0097
(branch af9015), as you expected the remote control doesn't react to
the generic USB_VID_TERRATEC matching and the given IR codes. No ill
effects though, the shipped remote just appears to be "dead" with those
settings. So these patches are fine and aren't a regression for my device.
I actually tried my luck with your suggestions for debugging the IR
support, but as the remote continued to act in a weird (bouncing keycodes
and no stable mappings at all), I didn't get very far. Except that the
EEPROM hash is 0x11f6768f.
A brutal first attempt to replace the predefined ir_codes_terratec/
af9015_ir_terratec with my current assumptions results in very similar,
erratic behaviour:
static struct ir_scancode ir_codes_terratec[] = {
{ 0x0057, KEY_1 },
{ 0x0020, KEY_2 },
{ 0x0026, KEY_3 },
{ 0x0056, KEY_4 },
{ 0x0021, KEY_5 },
{ 0x0027, KEY_6 },
{ 0x004b, KEY_7 },
{ 0x0022, KEY_8 },
{ 0x0009, KEY_9 },
{ 0x0023, KEY_0 },
{ 0x0024, KEY_CHANNELUP },
{ 0x0025, KEY_CHANNELDOWN },
{ 0x004e, KEY_ZOOM },
{ 0x0010, KEY_MUTE },
{ 0x001f, KEY_VOLUMEDOWN },
{ 0x001e, KEY_VOLUMEUP },
{ 0x001c, KEY_GOTO }, /* jump */
{ 0x043d, KEY_POWER },
};
static u8 af9015_ir_terratec[] = {
0x80, 0x7f, 0x12, 0xed, 0x3d, 0x04, 0x00, /* power */
0x80, 0x7f, 0x01, 0xfe, 0x10, 0x00, 0x00, /* mute */
0x80, 0x7f, 0x1a, 0xe5, 0x57, 0x00, 0x00, /* 1 */
0x80, 0x7f, 0x02, 0xfd, 0x56, 0x00, 0x00, /* 4 */
0x80, 0x7f, 0x1e, 0xe1, 0x4b, 0x00, 0x00, /* 7 */
0x80, 0x7f, 0x03, 0xfc, 0x4e, 0x00, 0x00, /* zoom */
0x80, 0x7f, 0x04, 0xfb, 0x1e, 0x00, 0x00, /* volume up */
0x80, 0x7f, 0x05, 0xfa, 0x1f, 0x00, 0x00, /* volume down */
0x80, 0x7f, 0x06, 0xf9, 0x20, 0x00, 0x00, /* 2 */
0x80, 0x7f, 0x07, 0xf8, 0x21, 0x00, 0x00, /* 5 */
0x80, 0x7f, 0x08, 0xf7, 0x22, 0x00, 0x00, /* 8 */
0x80, 0x7f, 0x09, 0xf6, 0x23, 0x00, 0x00, /* 0 */
0x80, 0x7f, 0x0a, 0xf5, 0x24, 0x00, 0x00, /* channel up */
0x80, 0x7f, 0x1b, 0xe4, 0x25, 0x00, 0x00, /* channel down */
0x80, 0x7f, 0x1f, 0xe0, 0x26, 0x00, 0x00, /* 3 */
0x80, 0x7f, 0x0d, 0xf2, 0x27, 0x00, 0x00, /* 6 */
0x80, 0x7f, 0x0c, 0xf3, 0x09, 0x00, 0x00, /* 9 */
0x80, 0x7f, 0x0e, 0xf1, 0x1c, 0x00, 0x00, /* jump */
};
The numbers seem to match, the rest was mostly "guessed" from the apparent
strategy (top-->down. left-->right), but pressing '1' still results in
"111111" (repeating until terminated by ctrl-c), pressing '4' in "54444445"
(until terminated).
Regards
Stefan Lippers-Hollmann
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [GIT PULL FOR 2.6.37] new AF9015 devices
2010-09-09 21:12 [GIT PULL FOR 2.6.37] new AF9015 devices Antti Palosaari
2010-09-09 23:19 ` Stefan Lippers-Hollmann
@ 2010-09-10 0:54 ` Stefan Lippers-Hollmann
2010-09-10 10:12 ` Antti Palosaari
2010-09-22 19:02 ` Mauro Carvalho Chehab
2 siblings, 1 reply; 6+ messages in thread
From: Stefan Lippers-Hollmann @ 2010-09-10 0:54 UTC (permalink / raw)
To: Antti Palosaari; +Cc: Mauro Carvalho Chehab, linux-media, TerraTux
Hi
On Friday 10 September 2010, Antti Palosaari wrote:
> Moikka Mauro!
> This patch series adds support for TerraTec Cinergy T Stick Dual RC and
> TerraTec Cinergy T Stick RC. Also MxL5007T devices with ref. design IDs
> should be working. Cinergy T Stick remote is most likely not working
> since it seems to use different remote as Cinergy T Dual... Stefan could
> you test and ensure T Stick is working?
>
> and thanks to TerraTec!
[...]
Another test and some further debugging of the IR core usedby the af9015
branch of this git tree led me to partial success. DVB-T functionality
continues to be fine and I've now found the proper values for this remote,
however once a key gets pressed, it is never released (unless another key
gets pressed which is then struck or unless I ctrl-c it (only works on
terminals). Likewise I'm not sure yet how to distinguish between the
"Cinergy T Dual" and my "Cinergy T RC MKII":
lsusb: 0ccd:0097 TerraTec Electronic GmbH Cinergy T RC MKII
EEPROM hash: 0x11f6768f
static struct ir_scancode ir_codes_terratec[] = {
{ 0x001e, KEY_1 },
{ 0x001f, KEY_2 },
{ 0x0020, KEY_3 },
{ 0x0021, KEY_4 },
{ 0x0022, KEY_5 },
{ 0x0023, KEY_6 },
{ 0x0024, KEY_7 },
{ 0x0025, KEY_8 },
{ 0x0026, KEY_9 },
{ 0x0027, KEY_0 },
{ 0x004b, KEY_CHANNELUP },
{ 0x004e, KEY_CHANNELDOWN },
{ 0x0009, KEY_ZOOM },
{ 0x0010, KEY_MUTE },
{ 0x0056, KEY_VOLUMEDOWN },
{ 0x0057, KEY_VOLUMEUP },
{ 0x001c, KEY_GOTO }, /* jump */
{ 0x043d, KEY_POWER },
};
static u8 af9015_ir_terratec[] = {
0x80, 0x7f, 0x12, 0xed, 0x3d, 0x04, 0x00, /* power */
0x80, 0x7f, 0x01, 0xfe, 0x10, 0x00, 0x00, /* mute */
0x80, 0x7f, 0x1a, 0xe5, 0x57, 0x00, 0x00, /* vol+ */
0x80, 0x7f, 0x02, 0xfd, 0x56, 0x00, 0x00, /* vol- */
0x80, 0x7f, 0x1e, 0xe1, 0x4b, 0x00, 0x00, /* ch+ */
0x80, 0x7f, 0x03, 0xfc, 0x4e, 0x00, 0x00, /* ch- */
0x80, 0x7f, 0x04, 0xfb, 0x1e, 0x00, 0x00, /* 1 */
0x80, 0x7f, 0x05, 0xfa, 0x1f, 0x00, 0x00, /* 2 */
0x80, 0x7f, 0x06, 0xf9, 0x20, 0x00, 0x00, /* 3 */
0x80, 0x7f, 0x07, 0xf8, 0x21, 0x00, 0x00, /* 4 */
0x80, 0x7f, 0x08, 0xf7, 0x22, 0x00, 0x00, /* 5 */
0x80, 0x7f, 0x09, 0xf6, 0x23, 0x00, 0x00, /* 6 */
0x80, 0x7f, 0x0a, 0xf5, 0x24, 0x00, 0x00, /* 7 */
0x80, 0x7f, 0x1b, 0xe4, 0x25, 0x00, 0x00, /* 8 */
0x80, 0x7f, 0x1f, 0xe0, 0x26, 0x00, 0x00, /* 9 */
0x80, 0x7f, 0x0d, 0xf2, 0x27, 0x00, 0x00, /* 0 */
0x80, 0x7f, 0x0c, 0xf3, 0x09, 0x00, 0x00, /* zoom */
0x80, 0x7f, 0x0e, 0xf1, 0x1c, 0x00, 0x00, /* jump */
};
Likewise lirc's irrecord comes up with this lircd.conf:
begin remote
name lircd.conf.Cinergy-T-RC-MKII
bits 56
eps 30
aeps 100
one 0 0
zero 0 0
pre_data_bits 8
pre_data 0x0
gap 452913
toggle_bit_mask 0x0
begin codes
KEY_1 0x01000200000001 0x00000000000000
KEY_2 0x01000300000001 0x00000000000000
KEY_3 0x01000400000001 0x00000000000000
KEY_4 0x01000500000001 0x00000000000000
KEY_5 0x01000600000001 0x00000000000000
KEY_6 0x01000700000001 0x00000000000000
KEY_7 0x01000800000001 0x00000000000000
KEY_8 0x01000900000001 0x00000000000000
KEY_9 0x01000A00000001 0x00000000000000
KEY_0 0x01000B00000001 0x00000000000000
KEY_CHANNELUP 0x01019200000001 0x00000000000000
KEY_CHANNELDOWN 0x01019300000001 0x00000000000000
KEY_ZOOM 0x01017400000001 0x00000000000000
KEY_MUTE 0x01007100000001 0x00000000000000
KEY_VOLUMEDOWN 0x01007200000001 0x00000000000000
KEY_VOLUMEUP 0x01007300000001 0x00000000000000
KEY_GOTO 0x01016200000001 0x00000000000000
KEY_POWER 0x01007400000001 0x00000000000000
end codes
end remote
Given that keys, once pressed, remain to be stuck, using both lirc's
dev/input and without any dæmon trying to catch keypresses, I have not
reached a functional configuration.
Regards
Stefan Lippers-Hollmann
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [GIT PULL FOR 2.6.37] new AF9015 devices
2010-09-10 0:54 ` Stefan Lippers-Hollmann
@ 2010-09-10 10:12 ` Antti Palosaari
2010-09-10 11:47 ` Stefan Lippers-Hollmann
0 siblings, 1 reply; 6+ messages in thread
From: Antti Palosaari @ 2010-09-10 10:12 UTC (permalink / raw)
To: Stefan Lippers-Hollmann; +Cc: linux-media
On 09/10/2010 03:54 AM, Stefan Lippers-Hollmann wrote:
> Another test and some further debugging of the IR core usedby the af9015
> branch of this git tree led me to partial success. DVB-T functionality
> continues to be fine and I've now found the proper values for this remote,
> however once a key gets pressed, it is never released (unless another key
> gets pressed which is then struck or unless I ctrl-c it (only works on
> terminals). Likewise I'm not sure yet how to distinguish between the
> "Cinergy T Dual" and my "Cinergy T RC MKII":
> Given that keys, once pressed, remain to be stuck, using both lirc's
> dev/input and without any dæmon trying to catch keypresses, I have not
> reached a functional configuration.
That`s known issue. Chip configures USB HID polling interval wrongly and
due to that HID starts repeating usually. There is USB-ID mapped quirks
in HID driver to avoid that, but only for few ADF9015 IDs...
I know how to fix that totally. I have been waiting new IR core merge
before switch remote from broken HID + polling to memory read based one.
But maybe I can do it just now and convert it later to IR core.
Antti
--
http://palosaari.fi/
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [GIT PULL FOR 2.6.37] new AF9015 devices
2010-09-10 10:12 ` Antti Palosaari
@ 2010-09-10 11:47 ` Stefan Lippers-Hollmann
0 siblings, 0 replies; 6+ messages in thread
From: Stefan Lippers-Hollmann @ 2010-09-10 11:47 UTC (permalink / raw)
To: Antti Palosaari; +Cc: linux-media
Hi
On Friday 10 September 2010, Antti Palosaari wrote:
> sure yet how to distinguish between the
> > "Cinergy T Dual" and my "Cinergy T RC MKII":
>
>
> > Given that keys, once pressed, remain to be stuck, using both lirc's
> > dev/input and without any dæmon trying to catch keypresses, I have not
> > reached a functional configuration.
>
> That`s known issue. Chip configures USB HID polling interval wrongly and
> due to that HID starts repeating usually. There is USB-ID mapped quir
Yes, now I see it. A quickly hacked test seems to improve the stuck key
events a lot (some keys still have a tendency to get stuck, but in general
it works):
--- a/drivers/hid/usbhid/hid-quirks.c
+++ b/drivers/hid/usbhid/hid-quirks.c
@@ -45,6 +45,7 @@ static const struct hid_blacklist {
{ USB_VENDOR_ID_TOPMAX, USB_DEVICE_ID_TOPMAX_COBRAPAD, HID_QUIRK_BADPAD },
{ USB_VENDOR_ID_AFATECH, USB_DEVICE_ID_AFATECH_AF9016, HID_QUIRK_FULLSPEED_INTERVAL },
+ { 0x0ccd, 0x0097, HID_QUIRK_FULLSPEED_INTERVAL },
{ USB_VENDOR_ID_ETURBOTOUCH, USB_DEVICE_ID_ETURBOTOUCH, HID_QUIRK_MULTI_INPUT },
{ USB_VENDOR_ID_PANTHERLORD, USB_DEVICE_ID_PANTHERLORD_TWIN_USB_JOYSTICK, HID_QUIRK_MULTI_INPUT | HID_QUIRK_SKIP_OUTPUT_REPORTS },
Regards
Stefan Lippers-Hollmann
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [GIT PULL FOR 2.6.37] new AF9015 devices
2010-09-09 21:12 [GIT PULL FOR 2.6.37] new AF9015 devices Antti Palosaari
2010-09-09 23:19 ` Stefan Lippers-Hollmann
2010-09-10 0:54 ` Stefan Lippers-Hollmann
@ 2010-09-22 19:02 ` Mauro Carvalho Chehab
2 siblings, 0 replies; 6+ messages in thread
From: Mauro Carvalho Chehab @ 2010-09-22 19:02 UTC (permalink / raw)
To: Antti Palosaari
Cc: Stefan Lippers-Hollmann, linux-media, TerraTux, Patrick Boettcher
Em 09-09-2010 18:12, Antti Palosaari escreveu:
> Moikka Mauro!
> This patch series adds support for TerraTec Cinergy T Stick Dual RC and TerraTec Cinergy T Stick RC. Also MxL5007T devices with ref. design IDs should be working. Cinergy T Stick remote is most likely not working since it seems to use different remote as Cinergy T Dual... Stefan could you test and ensure T Stick is working?
>
> and thanks to TerraTec!
>
> t. Antti
>
>
> The following changes since commit c9889354c6d36d6278ed851c74ace02d72efdd59:
>
> V4L/DVB: rc-core: increase repeat time (2010-09-08 13:04:40 -0300)
>
> are available in the git repository at:
> git://linuxtv.org/anttip/media_tree.git af9015
>
> Antti Palosaari (6):
> af9015: simple comment update
Hmm... dvb-usb.h defines it as:
struct dvb_usb_device_description devices[12];
It took me some time to find the current limit ;)
IMO, instead of just comment it as:
.num_device_descs = 9, /* check max from dvb-usb.h */
The better would be to add a definition at dvb-usb.h header for the max limit, and properly pointing it
on your drivers, like:
on dvb-usb.h:
#define MAX_DEVICES_PER_DEV_PROPS 12
...
struct dvb_usb_device_description devices[MAX_DEVICES_PER_DEV_PROPS];
on af9015 (and others):
.num_device_descs = 9, /* max is MAX_DEVICES_PER_DEV_PROPS as defined on dvb-usb.h */
I'll apply this patch to avoid breaking your series, but please provide me a fix.
> af9015: fix bug introduced by commit 490ade7e3f4474f626a8f5d778ead4e599b94fbcas
> af9013: add support for MaxLinear MxL5007T tuner
> af9015: add support for TerraTec Cinergy T Stick Dual RC
> af9015: add remote support for TerraTec Cinergy T Stick Dual RC
> af9015: map TerraTec Cinergy T Stick Dual RC remote to device ID
>
> drivers/media/dvb/dvb-usb/Kconfig | 1 +
> drivers/media/dvb/dvb-usb/af9015.c | 50 +++++++++++++----------
> drivers/media/dvb/dvb-usb/af9015.h | 63 +++++++++++++++++++++++++++++
> drivers/media/dvb/dvb-usb/dvb-usb-ids.h | 1 +
> drivers/media/dvb/frontends/af9013.c | 1 +
> drivers/media/dvb/frontends/af9013.h | 1 +
> drivers/media/dvb/frontends/af9013_priv.h | 5 +-
> 7 files changed, 99 insertions(+), 23 deletions(-)
>
>
Cheers,
Mauro
^ permalink raw reply [flat|nested] 6+ messages in thread