On Sat, Dec 21, 2024 at 01:26:25PM +0800, Pengyu Luo wrote: > On Sat, Dec 21, 2024 at 12:20 AM Johan Hovold <johan@xxxxxxxxxx> wrote: > > > On Sat, Dec 14, 2024 at 03:39:27PM +0200, Dmitry Baryshkov wrote: > > > On Sat, Dec 14, 2024 at 08:23:00PM +0800, Pengyu Luo wrote: > > > > > > >> + > > > > >> + /* /lib/firmware/ath11k/WCN6855/hw2.1/board-2.bin > > > > >> + * there is no calibrate data for huawei, > > > > >> + * but they have the same subsystem-device id > > > > >> + */ > > > > >> + qcom,ath11k-calibration-variant = "LE_X13S"; > > > > > > Finally, I found something, after I enabled ath11k boot dbg, I got my > > > > id_string='bus=pci,vendor=17cb,device=1103,subsystem-vendor=17cb,subsystem-device=0108,qmi-chip-id=2,qmi-board-id=255` > > > > > > > > With qca-swiss-army-knife (see [1]) > > > > > > > > $ ./ath11k-bdencoder -e board-2.bin | grep -i "$id_string" > > > > bus=pci,vendor=17cb,device=1103,subsystem-vendor=17cb,subsystem-device=0108,qmi-chip-id=2,qmi-board-id=255.bin created size: 60048 > > > > > > > > It have already been here. So that means I don't need to extract from > > > > Windows. I just extract it from linux-firmware then give it a variant > > > > name and send patches to ath11k, right? > > > > > > No. Usually 255 is an ID that is used by a variety of boards. So, > > > basically, you have to extract board data from Windows, add a proper > > > calibration variant that is specific to your board and then send the > > > resulting data to the ath11k mailing list. > > > > The board files used by Windows are not compatible with the Linux > > firmware, so the calibration data needs to come from Qualcomm. > > > > Then I don't understand why those bdwlan.* files would be in the > firmware tree. Those (Windows) files are supposedly used by the Windows driver firmware. The Linux firmware cannot consume those directly and Qualcomm had to provide us a corresponding file for the X13s (which they then included in board-2.bin and pushed to linux-firmware). Johan
