gp8psk: add systems supported by genpix devices to .delsys

---
 drivers/media/usb/dvb-usb/gp8psk-fe.c | 2 +-
 include/uapi/linux/dvb/frontend.h     | 4 ++++
 2 files changed, 5 insertions(+), 1 deletion(-)

index 223a3ca..fcdf82c 100644
@@ -333,7 +333,7 @@ success:
 
 
 static struct dvb_frontend_ops gp8psk_fe_ops = {
-	.delsys = { SYS_DVBS },
+	.delsys = { SYS_DCII_C_QPSK, SYS_DCII_I_QPSK, SYS_DCII_Q_QPSK, SYS_DCII_C_OQPSK, SYS_DSS, SYS_DVBS2, SYS_TURBO, SYS_DVBS },
 	.info = {
 		.name			= "Genpix DVB-S",
 		.frequency_min		= 800000,

index c56d77c..ada08a8 100644
@@ -410,6 +410,10 @@ typedef enum fe_delivery_system {
 	SYS_DVBT2,
 	SYS_TURBO,
 	SYS_DVBC_ANNEX_C,
+	SYS_DCII_C_QPSK,
+	SYS_DCII_I_QPSK,
+	SYS_DCII_Q_QPSK,
+	SYS_DCII_C_OQPSK,
 } fe_delivery_system_t;
 
 /* backward compatibility */

1.8.1.2

Genpix Skywalker and 8psk-to-usb devices do not support dvb-s2.

Correct, but many older userland applications used SYS_DVBS2 to tune before SYS_TURBO was added. I have no problem removing it but others might.

from gp8psk-fe.c

switch (c->delivery_system) {
case SYS_DVBS:
if (c->modulation != QPSK) {
deb_fe("%s: unsupported modulation selected (%d)\n", __func__, c->modulation);
return -EOPNOTSUPP;
}
c->fec_inner = FEC_AUTO;
break;
case SYS_DVBS2: /* kept for backwards compatibility */ deb_fe("%s: DVB-S2 delivery system selected\n", __func__); break;
case SYS_TURBO:
deb_fe("%s: Turbo-FEC delivery system selected\n", __func__); break;

Chris

I think the best solution here would be not to put false info in the driver and notify the author(s) of any apps still not updated to use SYS_TURBO, that they need to do so or let the communities for those apps fix it. Having the Genpix say it's dvb-s2 capable when it isn't can be a problem in systems with actual dvb-s2 sources which is why, iirc, SYS_TURBO was added in the first place. If nobody wants to bother fixing those apps (to my knowledge its only mythtv that has this problem) correctly and still wants to rely on the driver misrepresenting the devices capabilities then it seems appropriate that should be done in an external patch since SYS_TURBO already exists to prevent this.

Ive got no problem pulling it out, doesnt affect anything on my end. Do I need to resubmit the patch ?

Chris Lee