Hello, i encountered a bug! Steps to reproduce: 1. create GPT scheme and 1 partition on each drive 2. bcachefs format -L tank --data_replicas=2 --metadata_replicas=2 /dev/sdb1 /dev/sdc1 External UUID: 7e811c51-4281-424c-8682-cab29f6b683f Internal UUID: cef378e2-a2d6-497d-ae82-99e5fe4b699e Label: tank Version: 7 Block_size: 4.0K Btree node size: 256.0K Error action: remount-ro Clean: 0 Metadata replicas: have 0, want 2 Data replicas: have 0, want 2 Metadata checksum type: crc32c Data checksum type: crc32c Compression type: none String hash type: siphash 32 bit inodes: 0 GC reserve percentage: 8% Root reserve percentage: 0% Devices: 2 Device 0: UUID: 9231a3aa-8047-4072-aa2a-001ea6a2403f Size: 7.3T Bucket size: 256.0K First bucket: 0 Buckets: 30523537 Last mount: (never) State: readwrite Tier: 0 Has metadata: 0 Has data: 0 Replacement policy: lru Discard: 0 Device 1: UUID: c13809ec-1a69-4ba5-97cf-e2379d0d25f2 Size: 7.3T Bucket size: 256.0K First bucket: 0 Buckets: 30523537 Last mount: (never) State: readwrite Tier: 0 Has metadata: 0 Has data: 0 Replacement policy: lru Discard: 0 3. mount -t bcachefs /dev/sdb1:/dev/sdc1 /data 4. cd /data [root@nixos:/data]# dd if=/dev/zero of=testfile bs=1G count=1 oflag=direct The process is in status D (Uninterruptible sleep) https://en.wikipedia.org/wiki/Sleep_(system_call)#Uninterruptible_sleep dmesg: [Tue Jun 13 08:21:10 2017] bcachefs (7e811c51-4281-424c-8682-cab29f6b683f): initializing new filesystem [Tue Jun 13 08:24:13 2017] ------------[ cut here ]------------ [Tue Jun 13 08:24:13 2017] kernel BUG at /tmp/nix-build-linux-4.11-git.drv-0/bcachefs-43e3159/fs/bcachefs/buckets.c:236! [Tue Jun 13 08:24:13 2017] invalid opcode: 0000 [#1] SMP [Tue Jun 13 08:24:13 2017] Modules linked in: af_packet cfg80211 rfkill msr ipmi_ssif iTCO_wdt sb_edac edac_core x86_pkg_temp_thermal intel_powerclamp ast snd_hda_codec_realtek coretemp crct10dif_pclmul crc32_pclmul ghash_clmulni_intel snd_hda_codec_generic ttm pcbc drm_kms_helper ext4 aesni_intel drm aes_x86_64 crypto_simd glue_helper cryptd igb intel_cstate isci intel_uncore agpgart snd_hda_intel ptp fb_sys_fops firewire_ohci syscopyarea pps_core sysfillrect i2c_algo_bit sysimgblt crc16 firewire_core intel_rapl_perf snd_hda_codec libsas jbd2 crc_itu_t i2c_core evdev joydev ipmi_si mousedev input_leds led_class fscrypto mei_me ioatdma scsi_transport_sas mbcache snd_hda_core mac_hid mei lpc_ich shpchp snd_hwdep dca hed wmi ipmi_devintf ipmi_msghandler tpm_tis tpm_tis_core tpm button snd_pcm_oss snd_mixer_oss snd_pcm [Tue Jun 13 08:24:13 2017] snd_timer snd soundcore atkbd libps2 loop cpufreq_ondemand kvm_intel kvm irqbypass ip_tables x_tables ipv6 crc_ccitt autofs4 hid_generic usbhid hid bcachefs libcrc32c crc32c_generic sd_mod ahci libahci xhci_pci libata xhci_hcd ehci_pci ehci_hcd scsi_mod usbcore crc32c_intel usb_common rtc_cmos serio dm_mod [Tue Jun 13 08:24:13 2017] CPU: 0 PID: 14370 Comm: kworker/0:0H Not tainted 4.11.0 #1-NixOS [Tue Jun 13 08:24:13 2017] Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./EPC602D8A, BIOS P1.30 10/30/2014 [Tue Jun 13 08:24:13 2017] Workqueue: bcachefs bch2_write_index [bcachefs] [Tue Jun 13 08:24:13 2017] task: ffff89f42fe04100 task.stack: ffff99f28edb4000 [Tue Jun 13 08:24:13 2017] RIP: 0010:bch2_fs_usage_apply+0x196/0x1d0 [bcachefs] [Tue Jun 13 08:24:13 2017] RSP: 0000:ffff99f28edb7788 EFLAGS: 00010287 [Tue Jun 13 08:24:13 2017] RAX: 0000000000000100 RBX: ffff99f28edb7a80 RCX: 0000000000000000 [Tue Jun 13 08:24:13 2017] RDX: ffff89fa4d4181d8 RSI: ffff99f28edb7888 RDI: ffff89f76f6e0000 [Tue Jun 13 08:24:13 2017] RBP: ffff99f28edb7930 R08: 0000000000000000 R09: 0000000000000000 [Tue Jun 13 08:24:13 2017] R10: ffff89f434301010 R11: ffff89f434300000 R12: 0000000000001002 [Tue Jun 13 08:24:13 2017] R13: ffff89fa52597e00 R14: 0000000000001002 R15: ffff89fa52597e00 [Tue Jun 13 08:24:13 2017] FS: 0000000000000000(0000) GS:ffff89fa5f200000(0000) knlGS:0000000000000000 [Tue Jun 13 08:24:13 2017] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [Tue Jun 13 08:24:13 2017] CR2: ffffffffff600000 CR3: 0000000859560000 CR4: 00000000000406f0 [Tue Jun 13 08:24:13 2017] Call Trace: [Tue Jun 13 08:24:13 2017] ? bch2_insert_fixup_extent+0x1cb/0x760 [bcachefs] [Tue Jun 13 08:24:13 2017] __bch2_btree_insert_at+0x47c/0xcd0 [bcachefs] [Tue Jun 13 08:24:13 2017] ? bch2_inode_unpack+0x199/0x3f0 [bcachefs] [Tue Jun 13 08:24:13 2017] bchfs_write_index_update+0x322/0x530 [bcachefs] [Tue Jun 13 08:24:13 2017] ? kmem_cache_alloc+0xd7/0x1b0 [Tue Jun 13 08:24:13 2017] ? kernfs_path_from_node_locked+0x24e/0x350 [Tue Jun 13 08:24:13 2017] ? sg_init_table+0x1a/0x40 [Tue Jun 13 08:24:13 2017] ? blk_rq_map_sg+0x1ca/0x520 [Tue Jun 13 08:24:13 2017] ? scsi_init_sgtable+0x48/0x90 [scsi_mod] [Tue Jun 13 08:24:13 2017] ? bch2_readpages_end_io+0x70/0x70 [bcachefs] [Tue Jun 13 08:24:13 2017] ? dequeue_task_fair+0x566/0x1580 [Tue Jun 13 08:24:13 2017] ? pick_next_task_fair+0x3f4/0x450 [Tue Jun 13 08:24:13 2017] ? pick_next_task_fair+0x3f4/0x450 [Tue Jun 13 08:24:13 2017] ? __switch_to+0x161/0x4c0 [Tue Jun 13 08:24:13 2017] bch2_write_index+0x55/0x1f0 [bcachefs] [Tue Jun 13 08:24:13 2017] process_one_work+0x16b/0x4a0 [Tue Jun 13 08:24:13 2017] worker_thread+0x4b/0x510 [Tue Jun 13 08:24:13 2017] kthread+0x109/0x140 [Tue Jun 13 08:24:13 2017] ? pwq_unbound_release_workfn+0xe0/0xe0 [Tue Jun 13 08:24:13 2017] ? kthread_create_on_node+0x40/0x40 [Tue Jun 13 08:24:13 2017] ret_from_fork+0x2c/0x40 [Tue Jun 13 08:24:13 2017] Code: 7d d4 74 42 19 c9 83 c9 01 e9 5e ff ff ff 48 39 ce 74 21 48 39 f1 19 c9 83 c9 01 e9 4c ff ff ff 31 c9 48 39 c1 0f 8d 93 fe ff ff <0f> 0b f3 90 e9 ef fe ff ff 49 39 f8 74 14 4c 39 c7 19 c9 83 c9 [Tue Jun 13 08:24:13 2017] RIP: bch2_fs_usage_apply+0x196/0x1d0 [bcachefs] RSP: ffff99f28edb7788 [Tue Jun 13 08:24:13 2017] ---[ end trace 7b0d2186862d4fe8 ]— when i execute `mkdir /data/test` after that it is also in this state. rebooting also fails since it can’t get unmounted. after hard reboot i could mount the FS again. the file dd should create was not written. i was able to reproduce the error again. `dd if=/dev/zero of=/data/testfile3 bs=1G count=1` (without oflag=direct) works. kind regards davidak-- To unsubscribe from this list: send the line "unsubscribe linux-bcache" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
