When booting linux from kernels 5.15.41, 5.15.52, 5.18.12 and 5.19.4 (so I assume everything in between as well), the following message is displayed and bluetooth is not loaded: ``` [ 2.415827] usb 5-4: New USB device found, idVendor=0489, idProduct=e0cd, bcdDevice= 1.00 [ 2.415833] usb 5-4: New USB device strings: Mfr=5, Product=6, SerialNumber=7 [ 2.415835] usb 5-4: Product: Wireless_Device [ 2.415837] usb 5-4: Manufacturer: MediaTek Inc. [ 2.415838] usb 5-4: SerialNumber: 000000000 [ 2.436349] Loading firmware: mediatek/BT_RAM_CODE_MT7961_1_2_hdr.bin [ 2.436378] bluetooth hci0: Direct firmware load for mediatek/BT_RAM_CODE_MT7961_1_2_hdr.bin failed with error -2 [ 2.436384] Bluetooth: hci0: Failed to load firmware file (-2) [ 2.436387] Bluetooth: hci0: Failed to set up firmware (-2) ``` This results in a completely non-functional bluetooth, until a suspend/resume is performed, in which case around 80% of the time, the module will successfully load and bluetooth will work for the remainder of the session, until a second or further suspend/resume occurs, where I have observed another failure on one single occasion. If a suspend/resume does not resolve the issue, I find that it will stay broken on following suspend/resumes too. A reboot (not necessarily a shutdown, but simply using /sbin/reboot) is normally enough to get out of this dead loop. Kernel information: Linux version 5.19.4-gentoo-x86_64 (x86_64-pc-linux-gnu-gxx (Gentoo Hardened 11.3.0 p4) 11.3.0, GNU ld (Gentoo 2.38 p4) 2.38) #1 SMP PREEMPT_DYNAMIC Thu Aug 25 23:36:23 AEST 2022 I am running Linux on a Thinkpad L14 Gen 2 (AMD ver.) - the laptops DMI is reported as follows: ``` [ 0.000000] DMI: LENOVO 20X5004GAU/20X5004GAU, BIOS R1KET36W (1.21 ) 11/25/2021 ``` and kernel arguments: ``` [ 0.000000] Command line: BOOT_IMAGE=/vmlinuz-5.19.4-gentoo-x86_64 root=/dev/mapper/vg0-root ro zswap.enabled=1 zswap.compressor=lz4 dolvm crypt_root=UUID=xxxx_xxxx root_trim=yes ``` To trigger this problem, one must simply boot into Linux. Output of 'ver_linux': ``` Linux valentina 5.19.4-gentoo-x86_64 #1 SMP PREEMPT_DYNAMIC Thu Aug 25 23:36:23 AEST 2022 x86_64 AMD Ryzen 7 PRO 5850U with Radeon Graphics AuthenticAMD GNU/Linux GNU C 11.3.0 GNU Make 4.3 Binutils 2.38 Util-linux 2.37.4 Mount 2.37.4 Module-init-tools 29 E2fsprogs 1.46.5 Bison 3.8.2 Flex 2.6.4 Linux C++ Library 6.0.29 Dynamic linker (ldd) 2.35 Procps 3.3.17 Net-tools 2.10 Kbd 2.4.0 Console-tools 2.4.0 Sh-utils 8.32 Udev 250 Modules Loaded 8021q amdgpu asn1_encoder backlight ccm ccp cec cfg80211 cqhci crc32_pclmul crc32c_intel crc64 crc64_rocksoft crct10dif_pclmul des_generic dm_crypt dm_mod drm drm_display_helper drm_kms_helper drm_ttm_helper edac_mce_amd efi_pstore efivarfs ext4 fb_sys_fops firmware_attributes_class garp ghash_clmulni_intel gpu_sched i2c_algo_bit i2c_designware_core i2c_designware_platform i2c_scmi iommu_v2 iosf_mbi ip6_tables ip6t_rt ip6table_filter ip_tables ipt_REJECT iptable_filter jbd2 k10temp libarc4 libcrc32c libdes mac80211 mbcache md4 mfd_core mrp msr mt76 mt76_connac_lib mt7921_common mt7921e nf_conntrack nf_conntrack_broadcast nf_conntrack_ftp nf_conntrack_netbios_ns nf_defrag_ipv4 nf_defrag_ipv6 nf_log_syslog nf_nat nf_nat_ftp nf_reject_ipv4 nvme nvme_core nvram pkcs8_key_parser platform_profile rc_core roles sdhci sdhci_pci serio_raw sp5100_tco syscopyarea sysfillrect sysimgblt t10_pi think_lmi thinkpad_acpi trusted ttm typec_ucsi ucsi_acpi video wmi wmi_bmof xhci_hcd xhci_pci xhci_pci_renesas xt_LOG xt_addrtype xt_conntrack xt_hl xt_limit ``` I was hesitant to report this bug straight away as I saw there was some work being done on MediaTek drivers recently, but after compiling 5.19.4 and trying it for myself, the issue still persists. -- Finn Rayment
