[ Again, please don't top post. See http://en.wikipedia.org/wiki/Top_post ] On Sun, May 30, 2021 at 09:58:26PM +0300, Eero Lehtinen wrote: > Hi, > > I used dev_dbg instead of dev_info and got: > drivers/media/usb/dvb-usb-v2/rtl28xxu.c:648 > [dvb_usb_rtl28xxu]rtl28xxu_identify_state =_ "failed=%d\012" > drivers/media/usb/dvb-usb-v2/rtl28xxu.c:640 > [dvb_usb_rtl28xxu]rtl28xxu_identify_state =_ "chip_id=%u\012" > drivers/media/usb/dvb-usb-v2/rtl28xxu.c:630 > [dvb_usb_rtl28xxu]rtl28xxu_identify_state =_ "%s - ret2 = %d\012" > drivers/media/usb/dvb-usb-v2/rtl28xxu.c:627 > [dvb_usb_rtl28xxu]rtl28xxu_identify_state =_ "%s - ret1 = %d\012" > drivers/media/usb/dvb-usb-v2/rtl28xxu.c:620 > [dvb_usb_rtl28xxu]rtl28xxu_identify_state =_ "\012" > > dev_info does not show up with dmesg or dynamic debug. Odd. Just use dev_err() since that shows up in your logs. > Should the kernel use the CXD2837ER driver and not the mxl5005s driver > like it does with this patch. Yes, something like that could be the problem here. Did it use CXD2837ER before the patch? I still need to see the return values for those transfer to determine how best to address this so could you try again with dev_err()? Johan
