git: 40a4359064a5 - main - ctl_backend_block: Prepare for NVMe support
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Fri, 03 May 2024 00:16:05 UTC
The branch main has been updated by jhb: URL: https://cgit.FreeBSD.org/src/commit/?id=40a4359064a5d2db47d375a2ea7e251ee472b4e2 commit 40a4359064a5d2db47d375a2ea7e251ee472b4e2 Author: John Baldwin <jhb@FreeBSD.org> AuthorDate: 2024-05-02 23:34:04 +0000 Commit: John Baldwin <jhb@FreeBSD.org> CommitDate: 2024-05-02 23:38:30 +0000 ctl_backend_block: Prepare for NVMe support - Use wrapper routines for access to shared fields between SCSI and NVMe I/O requests. - Use protocol-agnostic wrapper routines for I/O completion status. Sponsored by: Chelsio Communications Differential Revision: https://reviews.freebsd.org/D44851 --- sys/cam/ctl/ctl_backend_block.c | 129 +++++++++++++++++++++++----------------- 1 file changed, 73 insertions(+), 56 deletions(-) diff --git a/sys/cam/ctl/ctl_backend_block.c b/sys/cam/ctl/ctl_backend_block.c index 56bc38cba07e..1cb833c3868c 100644 --- a/sys/cam/ctl/ctl_backend_block.c +++ b/sys/cam/ctl/ctl_backend_block.c @@ -390,6 +390,39 @@ ctl_complete_beio(struct ctl_be_block_io *beio) } } +static void +ctl_be_block_io_error(union ctl_io *io, int bio_cmd, uint16_t retry_count) +{ + switch (io->io_hdr.io_type) { + case CTL_IO_SCSI: + if (bio_cmd == BIO_FLUSH) { + /* XXX KDM is there is a better error here? */ + ctl_set_internal_failure(&io->scsiio, + /*sks_valid*/ 1, + retry_count); + } else { + ctl_set_medium_error(&io->scsiio, bio_cmd == BIO_READ); + } + break; + case CTL_IO_NVME: + switch (bio_cmd) { + case BIO_FLUSH: + case BIO_WRITE: + ctl_nvme_set_write_fault(&io->nvmeio); + break; + case BIO_READ: + ctl_nvme_set_unrecoverable_read_error(&io->nvmeio); + break; + default: + ctl_nvme_set_internal_error(&io->nvmeio); + break; + } + break; + default: + __assert_unreachable(); + } +} + static size_t cmp(uint8_t *a, uint8_t *b, size_t size) { @@ -408,7 +441,6 @@ ctl_be_block_compare(union ctl_io *io) struct ctl_be_block_io *beio; uint64_t off, res; int i; - uint8_t info[8]; beio = (struct ctl_be_block_io *)PRIV(io)->ptr; off = 0; @@ -421,15 +453,9 @@ ctl_be_block_compare(union ctl_io *io) break; } if (i < beio->num_segs) { - scsi_u64to8b(off, info); - ctl_set_sense(&io->scsiio, /*current_error*/ 1, - /*sense_key*/ SSD_KEY_MISCOMPARE, - /*asc*/ 0x1D, /*ascq*/ 0x00, - /*type*/ SSD_ELEM_INFO, - /*size*/ sizeof(info), /*data*/ &info, - /*type*/ SSD_ELEM_NONE); + ctl_io_set_compare_failure(io, off); } else - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); } static int @@ -442,7 +468,7 @@ ctl_be_block_move_done(union ctl_io *io, bool samethr) beio = (struct ctl_be_block_io *)PRIV(io)->ptr; DPRINTF("entered\n"); - io->scsiio.kern_rel_offset += io->scsiio.kern_data_len; + ctl_add_kern_rel_offset(io, ctl_kern_data_len(io)); /* * We set status at this point for read and compare commands. @@ -451,7 +477,7 @@ ctl_be_block_move_done(union ctl_io *io, bool samethr) (io->io_hdr.status & CTL_STATUS_MASK) == CTL_STATUS_NONE) { lbalen = ARGS(io); if (lbalen->flags & CTL_LLF_READ) { - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); } else if (lbalen->flags & CTL_LLF_COMPARE) { /* We have two data blocks ready for comparison. */ ctl_be_block_compare(io); @@ -545,19 +571,14 @@ ctl_be_block_biodone(struct bio *bio) error = beio->first_error; if (error != 0) { if (error == EOPNOTSUPP) { - ctl_set_invalid_opcode(&io->scsiio); + ctl_io_set_invalid_opcode(io); } else if (error == ENOSPC || error == EDQUOT) { - ctl_set_space_alloc_fail(&io->scsiio); + ctl_io_set_space_alloc_fail(io); } else if (error == EROFS || error == EACCES) { - ctl_set_hw_write_protected(&io->scsiio); - } else if (beio->bio_cmd == BIO_FLUSH) { - /* XXX KDM is there is a better error here? */ - ctl_set_internal_failure(&io->scsiio, - /*sks_valid*/ 1, - /*retry_count*/ 0xbad2); + ctl_io_set_hw_write_protected(io); } else { - ctl_set_medium_error(&io->scsiio, - beio->bio_cmd == BIO_READ); + ctl_be_block_io_error(io, beio->bio_cmd, + /*retry_count*/ 0xbad2); } ctl_complete_beio(beio); return; @@ -571,12 +592,12 @@ ctl_be_block_biodone(struct bio *bio) || (beio->bio_cmd == BIO_FLUSH) || (beio->bio_cmd == BIO_DELETE) || (ARGS(io)->flags & CTL_LLF_VERIFY)) { - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); ctl_complete_beio(beio); } else { if ((ARGS(io)->flags & CTL_LLF_READ) && beio->beio_cont == NULL) { - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); if (cbe_lun->serseq >= CTL_LUN_SERSEQ_SOFT) ctl_serseq_done(io); } @@ -614,12 +635,10 @@ ctl_be_block_flush_file(struct ctl_be_block_lun *be_lun, mtx_unlock(&be_lun->io_lock); if (error == 0) - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); else { - /* XXX KDM is there is a better error here? */ - ctl_set_internal_failure(&io->scsiio, - /*sks_valid*/ 1, - /*retry_count*/ 0xbad1); + ctl_be_block_io_error(io, BIO_FLUSH, + /*retry_count*/ 0xbad1); } ctl_complete_beio(beio); @@ -704,7 +723,7 @@ ctl_be_block_dispatch_file(struct ctl_be_block_lun *be_lun, SDT_PROBE0(cbb, , read, file_done); if (error == 0 && xuio.uio_resid > 0) { /* - * If we red less then requested (EOF), then + * If we read less then requested (EOF), then * we should clean the rest of the buffer. */ s = beio->io_len - xuio.uio_resid; @@ -759,12 +778,11 @@ ctl_be_block_dispatch_file(struct ctl_be_block_lun *be_lun, */ if (error != 0) { if (error == ENOSPC || error == EDQUOT) { - ctl_set_space_alloc_fail(&io->scsiio); + ctl_io_set_space_alloc_fail(io); } else if (error == EROFS || error == EACCES) { - ctl_set_hw_write_protected(&io->scsiio); + ctl_io_set_hw_write_protected(io); } else { - ctl_set_medium_error(&io->scsiio, - beio->bio_cmd == BIO_READ); + ctl_be_block_io_error(io, beio->bio_cmd, 0); } ctl_complete_beio(beio); return; @@ -776,12 +794,12 @@ ctl_be_block_dispatch_file(struct ctl_be_block_lun *be_lun, */ if ((beio->bio_cmd == BIO_WRITE) || (ARGS(io)->flags & CTL_LLF_VERIFY)) { - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); ctl_complete_beio(beio); } else { if ((ARGS(io)->flags & CTL_LLF_READ) && beio->beio_cont == NULL) { - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); if (cbe_lun->serseq > CTL_LUN_SERSEQ_SOFT) ctl_serseq_done(io); } @@ -918,18 +936,18 @@ ctl_be_block_unmap_file(struct ctl_be_block_lun *be_lun, */ switch (error) { case 0: - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); break; case ENOSPC: case EDQUOT: - ctl_set_space_alloc_fail(&io->scsiio); + ctl_io_set_space_alloc_fail(io); break; case EROFS: case EACCES: - ctl_set_hw_write_protected(&io->scsiio); + ctl_io_set_hw_write_protected(io); break; default: - ctl_set_medium_error(&io->scsiio, false); + ctl_be_block_io_error(io, BIO_DELETE, 0); } ctl_complete_beio(beio); } @@ -1008,12 +1026,11 @@ ctl_be_block_dispatch_zvol(struct ctl_be_block_lun *be_lun, */ if (error != 0) { if (error == ENOSPC || error == EDQUOT) { - ctl_set_space_alloc_fail(&io->scsiio); + ctl_io_set_space_alloc_fail(io); } else if (error == EROFS || error == EACCES) { - ctl_set_hw_write_protected(&io->scsiio); + ctl_io_set_hw_write_protected(io); } else { - ctl_set_medium_error(&io->scsiio, - beio->bio_cmd == BIO_READ); + ctl_be_block_io_error(io, beio->bio_cmd, 0); } ctl_complete_beio(beio); return; @@ -1025,12 +1042,12 @@ ctl_be_block_dispatch_zvol(struct ctl_be_block_lun *be_lun, */ if ((beio->bio_cmd == BIO_WRITE) || (ARGS(io)->flags & CTL_LLF_VERIFY)) { - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); ctl_complete_beio(beio); } else { if ((ARGS(io)->flags & CTL_LLF_READ) && beio->beio_cont == NULL) { - ctl_set_success(&io->scsiio); + ctl_io_set_success(io); if (cbe_lun->serseq > CTL_LUN_SERSEQ_SOFT) ctl_serseq_done(io); } @@ -1700,16 +1717,16 @@ ctl_be_block_dispatch(struct ctl_be_block_lun *be_lun, } if (bptrlen->len < lbalen->len) beio->beio_cont = ctl_be_block_next; - io->scsiio.be_move_done = ctl_be_block_move_done; + ctl_set_be_move_done(io, ctl_be_block_move_done); /* For compare we have separate S/G lists for read and datamove. */ if (beio->two_sglists) - io->scsiio.kern_data_ptr = (uint8_t *)&beio->sg_segs[CTLBLK_HALF_SEGS]; + ctl_set_kern_data_ptr(io, &beio->sg_segs[CTLBLK_HALF_SEGS]); else - io->scsiio.kern_data_ptr = (uint8_t *)beio->sg_segs; - io->scsiio.kern_data_len = beio->io_len; - io->scsiio.kern_sg_entries = beio->num_segs; - io->scsiio.kern_data_ref = ctl_refcnt_beio; - io->scsiio.kern_data_arg = beio; + ctl_set_kern_data_ptr(io, beio->sg_segs); + ctl_set_kern_data_len(io, beio->io_len); + ctl_set_kern_sg_entries(io, beio->num_segs); + ctl_set_kern_data_ref(io, ctl_refcnt_beio); + ctl_set_kern_data_arg(io, beio); io->io_hdr.flags |= CTL_FLAG_ALLOCATED; /* @@ -1749,7 +1766,7 @@ ctl_be_block_worker(void *context, int pending) mtx_unlock(&be_lun->queue_lock); beio = (struct ctl_be_block_io *)PRIV(io)->ptr; if (cbe_lun->flags & CTL_LUN_FLAG_NO_MEDIA) { - ctl_set_busy(&io->scsiio); + ctl_io_set_busy(io); ctl_complete_beio(beio); continue; } @@ -1762,7 +1779,7 @@ ctl_be_block_worker(void *context, int pending) STAILQ_REMOVE_HEAD(&be_lun->config_write_queue, links); mtx_unlock(&be_lun->queue_lock); if (cbe_lun->flags & CTL_LUN_FLAG_NO_MEDIA) { - ctl_set_busy(&io->scsiio); + ctl_io_set_busy(io); ctl_config_write_done(io); continue; } @@ -1775,7 +1792,7 @@ ctl_be_block_worker(void *context, int pending) STAILQ_REMOVE_HEAD(&be_lun->config_read_queue, links); mtx_unlock(&be_lun->queue_lock); if (cbe_lun->flags & CTL_LUN_FLAG_NO_MEDIA) { - ctl_set_busy(&io->scsiio); + ctl_io_set_busy(io); ctl_config_read_done(io); continue; } @@ -1788,7 +1805,7 @@ ctl_be_block_worker(void *context, int pending) STAILQ_REMOVE_HEAD(&be_lun->input_queue, links); mtx_unlock(&be_lun->queue_lock); if (cbe_lun->flags & CTL_LUN_FLAG_NO_MEDIA) { - ctl_set_busy(&io->scsiio); + ctl_io_set_busy(io); ctl_data_submit_done(io); continue; }