)]}' { "log": [ { "commit": "a026757ff56365b4aa3875c14f1bd5733e0e8bb2", "tree": "c9cd1990da5bc8cc11808a78067e96923091e006", "parents": [ "38b11bae6ba02da352340aff12ee25755977b222" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Oct 01 17:23:22 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Oct 02 14:16:08 2012 -0700" }, "message": "target: Add target_submit_cmd_map_sgls for SGL fabric memory passthrough\n\nThis patch adds a new target_submit_cmd_map_sgls() to pass pre-allocated\nSGL memory using transport_generic_map_mem_to_cmd() logic into the generic\ntarget submit I/O codepath.\n\nIt also adds a target_submit_cmd() wrapper around target_submit_cmd_map_sgls()\nfor existing fabric code that already assumes internal target-core SGL memory\nallocation.\n\n(v2: Rename to target_submit_cmd_map_sgls + drop TARGET_SCF_MAP_MEM flag\n in favor of non zero sgl_count check)\n\nReported-by: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "fd30e9310253235e88a4eaa38fa92439b56c220e", "tree": "eaa3c34f2c0baf84a407c92ff9d1f05b273e74fb", "parents": [ "93d441a888c4aefd9f37b30911f8fefa27bdfda3" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun Aug 26 13:35:58 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Sep 17 17:13:34 2012 -0700" }, "message": "target: Drop se_subsystem_api-\u003e[write_cache,fua_write]_emulated flags\n\nThis patch drops se_subsystem_api-\u003e[write_cache,fua_write]_emulated flags\nset by viritual FILEIO/IBLOCK/RD_MCP backend drivers in favor of explict\nTRANSPORT_PLUGIN_PHBA_PDEV checks to know when to fail if userspace is\nattempting to set virtual emulation bits for an pSCSI (passthrough)\nbackend device.\n\nReported-by: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "9c58b7ddd70dd7bfaac4ca87131f36d10aaba441", "tree": "2796f11c283904cef9b38f543e31fed3fea565c7", "parents": [ "2ed772b7b9df0f459308b3cbececc0136076d09e" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Aug 15 14:35:25 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Sep 17 17:12:58 2012 -0700" }, "message": "target: Simplify fabric sense data length handling\n\nEvery fabric driver has to supply a se_tfo-\u003eset_fabric_sense_len()\nmethod, just so iSCSI can return an offset of 2. However, every fabric\ndriver is already allocating a sense buffer and passing it into the\ntarget core, either via transport_init_se_cmd() or target_submit_cmd().\n\nSo instead of having iSCSI pass the start of its sense buffer into the\ncore and then later tell the core to skip the first 2 bytes, it seems\neasier for iSCSI just to do the offset of 2 when it passes the sense\nbuffer into the core. Then we can drop the se_tfo-\u003eset_fabric_sense_len()\neverywhere, and just add a couple of lines of code to iSCSI to set the\nsense data length to the beginning of the buffer right before it sends\nit over the network.\n\n(nab: Remove .set_fabric_sense_len usage from tcm_qla2xxx_npiv_ops +\n change transport_get_sense_buffer to follow v3.6-rc6 code w/o\n -\u003eset_fabric_sense_len usage)\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "2ed772b7b9df0f459308b3cbececc0136076d09e", "tree": "4f97f1cbdb889b921d48799c085e5142e5aa001b", "parents": [ "343d475d6176fa081918fcbf3fcb0b0143e28661" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Aug 15 14:35:24 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Sep 17 16:15:47 2012 -0700" }, "message": "target: Remove unused target_core_fabric_ops.get_fabric_sense_len method\n\nThere are no callers of se_tfo-\u003eget_fabric_sense_len(), so we should\nstop having every fabric driver implement it.\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "32a8811ff164f882712c17946e58e52444f464a7", "tree": "f7af9ff41cd4f4e6c61a831c4e14fb1d57ebe05d", "parents": [ "3717ef0c63e90686d959158e9728a13a49229be6" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Fri Sep 07 17:30:36 2012 +0200" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Sep 07 11:14:21 2012 -0700" }, "message": "target: support zero allocation length in REQUEST SENSE\n\nSimilar to INQUIRY and MODE SENSE, construct the sense data in a\nbuffer and later copy it to the scatterlist. Do not do anything,\nbut still clear a pending unit attention condition, if the allocation\nlength is zero.\n\nHowever, SPC tells us that \"If a REQUEST SENSE command is terminated with\nCHECK CONDITION status [and] the REQUEST SENSE command was received on\nan I_T nexus with a pending unit attention condition (i.e., before the\ndevice server reports CHECK CONDITION status), then the device server\nshall not clear the pending unit attention condition.\" Do the\ntransport_kmap_data_sg early to detect this case.\n\nIt also tells us \"Device servers shall not adjust the additional sense\nlength to reflect truncation if the allocation length is less than the\nsense data available\", so do not do that! Note that the err variable\nis write-only.\n\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "d5829eac5f7cfff89c6d1cf11717eee97cf030d0", "tree": "9acff1b99c654235b5ad4534735fdaf03a9c5a45", "parents": [ "27a2709912ac19c755d34c79fe11994b0bf8082b" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Wed Sep 05 17:09:15 2012 +0200" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Sep 05 17:20:28 2012 -0700" }, "message": "target: fix use-after-free with PSCSI sense data\n\nThe pointer to the sense buffer is fetched by transport_get_sense_data,\nbut this is called by target_complete_ok_work long after pscsi_req_done\nhas freed the struct that contains it.\n\nPass instead the fabric\u0027s sense buffer to transport_complete,\nand copy the data to it directly in transport_complete. Setting\nSCF_TRANSPORT_TASK_SENSE also becomes a duty of transport_complete.\n\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "af74115eed22698f771fec1287a864975c9a6671", "tree": "e0619a894a047598aaa412b334c06e20f2eb8fe5", "parents": [ "08a16208c8cb2ce1f79fea24f21dd7a8df4f12b6" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Aug 15 21:24:52 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Aug 21 14:27:35 2012 -0700" }, "message": "target: Remove unused se_cmd.cmd_spdtl\n\nThis was originally for helping fabrics to determine overflow/underflow\nstatus, and has been superceeded by SCF_OVERFLOW_BIT + SCF_UNDERFLOW_BIT.\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "d6dfc868bcf329392abd1ecfa7357eb51ebf8c30", "tree": "7eb33785e4d6a59111560b56e68f4facaddee310", "parents": [ "7409a6657aebf8be74c21d0eded80709b27275cb" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Jul 16 11:04:39 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Jul 17 17:05:05 2012 -0700" }, "message": "target: Allow for target_submit_cmd() returning errors\n\nWe want it to be possible for target_submit_cmd() to return errors up\nto its fabric module callers. For now just update the prototype to\nreturn an int, and update all callers to handle non-zero return values\nas an error.\n\nThis is immediately useful for tcm_qla2xxx to fix a long-standing active\nI/O session shutdown race, but tcm_fc, usb-gadget, and sbp-target the\nfabric maintainers need to check + ACK that handling a target_submit_cmd()\nfailure due to session shutdown does not introduce regressions\n\n(nab: Respin against for-next after initial NACK + update docbook comment +\n fix double se_cmd init in exception path for usb-gadget)\n\nCc: Chad Dupuis \u003cchad.dupuis@qlogic.com\u003e\nCc: Arun Easi \u003carun.easi@qlogic.com\u003e\nCc: Chris Boot \u003cbootc@bootc.net\u003e\nCc: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\nCc: Mark Rustad \u003cmark.d.rustad@intel.com\u003e\nCc: Sebastian Andrzej Siewior \u003cbigeasy@linutronix.de\u003e\nCc: Felipe Balbi \u003cbalbi@ti.com\u003e\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "e2397c704429025bc6b331a970f699e52f34283e", "tree": "470157d2f9a66ca40b295b9aa16c4fdf2222f15c", "parents": [ "c8045372dd2d4784bc4cb32ddc27f4b88ff2140c" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Jul 16 15:34:21 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:35 2012 -0700" }, "message": "target: Add generation of LOGICAL BLOCK ADDRESS OUT OF RANGE\n\nMany SCSI commands are defined to return a CHECK CONDITION / ILLEGAL\nREQUEST with ASC set to LOGICAL BLOCK ADDRESS OUT OF RANGE if the\ninitiator sends a command that accesses a too-big LBA. Add an enum\nvalue and case entries so that target code can return this status.\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "1c7b13fe65269960f63082eafccede547191ab02", "tree": "2c2d7f46ab88c31f3b347394ef0d51eb9d11ab8c", "parents": [ "e1013f14376f3121e73917f5455b9b7a22bdfb41" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Jul 16 11:04:42 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:27 2012 -0700" }, "message": "target: Remove se_session.sess_wait_list\n\nSince we set se_session.sess_tearing_down and stop new commands from\nbeing added to se_session.sess_cmd_list before we wait for commands to\nfinish when freeing a session, there\u0027s no need for a separate\nsess_wait_list -- if we let new commands be added to sess_cmd_list\nafter setting sess_tearing_down, that would be a bug that breaks the\nlogic of waiting in-flight commands.\n\nAlso rename target_splice_sess_cmd_list() to\ntarget_sess_cmd_list_set_waiting(), since we are no longer splicing\nonto a separate list.\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "669ab62c9d045bd2ff647f39e9e7a088e7e8706a", "tree": "b4b10367aa41167ae48b00b9589abacc076987b8", "parents": [ "b46e34a6721cb057702f19ebe27cfc7877e500e5" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Jul 16 11:04:37 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:25 2012 -0700" }, "message": "target: Un-export target_get_sess_cmd()\n\nThere are no in-tree users of target_get_sess_cmd() outside of\ntarget_core_transport.c. Any new code should use the higher-level\ntarget_submit_cmd() interface. So let\u0027s un-export target_get_sess_cmd()\nand make it static to the one file where it\u0027s actually used.\n\n(nab: Fix up minor fuzz to for-next)\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "af8772926f019b7bddd7477b8de5f3b0f12bad21", "tree": "99bdfc7335a3223985b033404fb06fc58cb754f0", "parents": [ "f314643751450a582c1ca40a54558240ef7cd4bf" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jul 08 15:58:49 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:21 2012 -0700" }, "message": "target: replace the processing thread with a TMR work queue\n\nThe last functionality of the target processing thread is offloading possibly\nlong running task management requests from the submitter context. To keep\nTMR semantics the same we need a single threaded ordered queue, which can\nbe provided by a per-device workqueue with the right flags.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "f314643751450a582c1ca40a54558240ef7cd4bf", "tree": "482343a17d77d36a872abff047bac5406728fae0", "parents": [ "7a6f0a1ea5f5ac9b67211071d558c5b09580695d" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jul 08 15:58:48 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:21 2012 -0700" }, "message": "target: remove transport_generic_handle_cdb_map\n\nRemove this command submission path which is not used by any in-tree driver.\nThis also removes the now unused new_cmd_map fabtric method, which a few\ndrivers implemented despite never calling transport_generic_handle_cdb_map.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "1389533ef944823a6ebc170345ad8743e48bc404", "tree": "53cb646e75d379521e205175505c4d9939b4589b", "parents": [ "b8b22533fed12dbb9e5a63d414cb1c768d1c28dd" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jul 08 15:58:46 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:20 2012 -0700" }, "message": "target: remove transport_generic_handle_data\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "70baf0ab3b2608727515086bee4c484a93e22880", "tree": "9cb6a054b5b83c44629e3e1408839b570a228658", "parents": [ "f7113a47e2c3880ed5f8a05ff9c54ead4ee7351c" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jul 08 15:58:39 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:16 2012 -0700" }, "message": "target: remove transport_generic_process_write\n\nJust call target_execute_cmd directly. Also, convert loopback, sbp,\nusb-gadget to use the newly exported target_execute_cmd().\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "14150a6bbe9e15ce8e7a4f79047c2b4284a51b3d", "tree": "1a77a526359e6159ed367ac3b1e8f1e13de6bb37", "parents": [ "6f974e8ce7b3f661910a49c7c2ba095631f341e9" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jun 17 18:40:55 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:35:14 2012 -0700" }, "message": "target: move unmap to struct spc_ops\n\nHaving all the unmap payload parsing in the backed is a bit ugly, but until\nmore drivers support it and we can find a good interface for all of them\nthat seems the way to go.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "6f974e8ce7b3f661910a49c7c2ba095631f341e9", "tree": "5698f1ed849b92eec547571594a68183c592b590", "parents": [ "ad67f0d9e63ca94661e06a145f05a9302368a826" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jun 17 18:40:54 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:34:41 2012 -0700" }, "message": "target: move write_same to struct spc_ops\n\nAdd spc_ops-\u003eexecute_write_same() caller for -\u003eexecute_cmd() setup,\nand update IBLOCK backends to use it.\n\n(nab: add export of spc_get_write_same_sectors symbol)\n(roland: Carry forward: Fix range calculation in WRITE SAME emulation\n when num blocks \u003d\u003d 0)\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "ad67f0d9e63ca94661e06a145f05a9302368a826", "tree": "2ada5604a0f61fa1f933f29cd4674f0cb95595c6", "parents": [ "0c2ad7d1132d8b089b1d37875917858e03610019" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jun 17 18:40:53 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:29:12 2012 -0700" }, "message": "target: move sync_cache to struct spc_ops\n\nAdd spc_ops-\u003eexecute_sync_cache() caller for -\u003eexecute_cmd() setup,\nand update IBLOCK + FILEIO backends to use it.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "0c2ad7d1132d8b089b1d37875917858e03610019", "tree": "e86cc118497605026a15c0e97630c57aeb625793", "parents": [ "e1306bdab3af8bef620990a99e613f99eb30065c" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jun 17 18:40:52 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:29:12 2012 -0700" }, "message": "target: add struct spc_ops + initial -\u003eexecute_rw pointer usage\n\nRemove the execute_cmd method in struct se_subsystem_api, and always use the\none directly in struct se_cmd. To make life simpler for SBC virtual backends\na struct spc_ops that is passed to sbc_parse_cmd is added. For now it\nonly contains an execute_rw member, but more will follow with the subsequent\ncommits.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "e1306bdab3af8bef620990a99e613f99eb30065c", "tree": "8712e8e9777bdc59f44f6801decf94d077c33135", "parents": [ "c75660c5e42c1a3edc4df0d062b2bd73e69dadb7" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun Jun 17 18:40:51 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:29:12 2012 -0700" }, "message": "target: remove dead SCF_ flags\n\nRemove the dead SCF_SE_ALLOW_EOO and SCF_DELAYED_CMD_FROM_SAM_ATTR\nfrom se_cmd_flags_table.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "9f3eb93eaf5bd49051458fcb098af3fcf111d3d0", "tree": "449b9eafb1921bbfce0684e15b098691ddc2e412", "parents": [ "5f41a31d0a49a014adb1588edd0cc7f7e30cc55b" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun May 20 14:35:02 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:29:11 2012 -0700" }, "message": "target: move ref_cmd from the generic se_tmr_req into iscsi code\n\nAlso remove the unused ref_task_lun field in struct se_tmr_req.\n\n(nab: Add missing TASK_REASSIGN ref_lun vs. ref_cmd orig_fe_lun checks\n in iscsit_tmr_task_reassign)\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "5f41a31d0a49a014adb1588edd0cc7f7e30cc55b", "tree": "6ee74b84c59a0473aba1257d2e762fa6935c4066", "parents": [ "a3785c8740c5b56b49ec336b59be996393d83332" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun May 20 14:34:44 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:29:11 2012 -0700" }, "message": "target: remove the execute list\n\nSince \"target: Drop se_device TCQ queue_depth usage from I/O path\" we always\nsubmit all commands (or back then, tasks) from __transport_execute_tasks.\n\nThat means the the execute list has lots its purpose, as we can simply\nsubmit the commands that are restarted in transport_complete_task_attr\ndirectly while we walk the list. In fact doing so also solves a race\nin the way it currently walks to delayed_cmd_list as well.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "1fd032ee10d2816c947f5d5b9abda95e728f0a8f", "tree": "2a76ad6b466160127dd41ba1d2f6fdb54f412bf2", "parents": [ "d6e0175cf3f9737a760482d185bb73566bcc9331" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun May 20 11:59:15 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:27:45 2012 -0700" }, "message": "target: move code for CDB emulation\n\nMove the existing code in target_core_cdb.c into the files for the command\nsets that the emulations implement.\n\n(roland + nab: Squash patch: Fix range calculation in WRITE SAME emulation\n when num blocks \u003d\u003d 0s)\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "d6e0175cf3f9737a760482d185bb73566bcc9331", "tree": "99ca8361ebc8c4e83f8335193963faa3f30007b5", "parents": [ "88455ec4be02c395820b1ff57656b0844ec03ac3" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun May 20 11:59:14 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:25:56 2012 -0700" }, "message": "target: add a parse_cdb method to the backend drivers\n\nInstead of trying to handle all SCSI command sets in one function\n(transport_generic_cmd_sequencer) call out to the backend driver to perform\nthis functionality. For pSCSI a copy of the existing code is used, but for\nall virtual backends we can use a new parse_sbc_cdb helper is used to\nprovide a simple SBC emulation.\n\nFor now this setups means a fair amount of duplication between pSCSI and the\nSBC library, but patches later in this series will sort out that problem.\n\n(nab: Fix up build failure in target_core_pscsi.c)\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "64f1db38c65fa634f4aa21e0f70480a6b8b4d47c", "tree": "de8a6b08d9230ba7cb69a3c78174b351123747df", "parents": [ "cb4f4d3c7398a709b48d397e0520ee2509a953a4" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Sun May 20 11:59:11 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Jul 16 17:25:55 2012 -0700" }, "message": "target: remove control CDB flags\n\nWe don\u0027t need three flags to classifiy the CDB as we can check for a NULL S/G\nlist for a dataless command, and can infer from the absence of the data flag\nthat we deal with a control CDB. Also remove the _SG_IO from the data CDB\nflag as all I/O is dont on S/G lists now.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "4149268e7816d719b0fde8e89aaa6db8c168fc43", "tree": "3c180fca14317541addad2cf65e0d5ad5b01bc3d", "parents": [ "cfaf025112d3856637ff34a767ef785ef5cf2ca9" ], "author": { "name": "Joern Engel", "email": "joern@logfs.org", "time": "Fri May 18 13:57:19 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Jun 12 19:33:05 2012 -0700" }, "message": "target: Add TFO-\u003eput_session() caller for HW fabric session shutdown\n\nThis patch adds an optional target_core_fabric_ops-\u003eput_session() caller\nwithin the existing target_put_session() code path.\n\nThis is required by tcm_qla2xxx code in order to invoke it\u0027s own fabric\nspecific session shutdown handler using se_session-\u003esess_kref.\n\nSigned-off-by: Joern Engel \u003cjoern@logfs.org\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Arun Easi \u003carun.easi@qlogic.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "5b9a4d7280e160982a8ea37bc03619f53b5c98b7", "tree": "3d7fbce10e7379097f34276b6a8a7d090d4dc813", "parents": [ "ba539743b70cd160c84bab1c82910d0789b820f8" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed May 16 22:02:34 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Thu May 17 00:45:58 2012 -0700" }, "message": "target: Add MI_REPORT_TARGET_PGS ext. header + implict_trans_secs attribute\n\nThis patch adds support for ALUA MI_REPORT_TARGET_PGS extended header\nformat defined within SPC-4. It changes target core ALUA emulation logic\nwithin target_emulate_report_target_port_groups() to support both the\nextended and original length only header formats.\n\nIt includes adding a new \u0027implict_trans_secs\u0027 attribute for each ALUA\ntarget port group to control the value returned to the application client\nfor an recommended implict translation timeout in seconds. By default\nthis value is currently set to zero, and limited up to 255 by virtue of\nusing a single byte in the extended header format.\n\nThis value is used by target_emulate_report_target_port_groups() within\nthe extended header logic to set IMPLICIT TRANSITION TIME as defined by\nspc4r30.\n\nCc: Hannes Reinecke \u003chare@suse.de\u003e\nCc: Rob Evers \u003crevers@redhat.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "33933a0e962bba4af32de7db912e2fdfef603d24", "tree": "df8e63fc5554c5eae2950f9d07658e5bc3ce6b64", "parents": [ "d28d96dcf635410422b41eca5203b0470dd059c1" ], "author": { "name": "Jörn Engel", "email": "joern@logfs.org", "time": "Fri May 11 10:35:08 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri May 11 15:01:52 2012 -0700" }, "message": "target: make target_put_session void\n\nNo real change, it effectively already was.\n\nSigned-off-by: Joern Engel \u003cjoern@logfs.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "11e764bd5ed4bb930e0ec5dd161df58307507347", "tree": "fa753a950d19c2ef848c50c60f71019dcb912e74", "parents": [ "2301917044b96fda41f794011368e623a9b7a435" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed May 09 12:42:09 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed May 09 15:08:47 2012 -0700" }, "message": "target: Remove max_sectors device attribute for modern se_task less code\n\nThis patch removes the original usage of dev_attr-\u003emax_sectors in favor of\ndev_attr-\u003ehw_max_sectors that is now being enforced by target core from\nwithin transport_generic_cmd_sequencer() for SCF_SCSI_DATA_SG_IO_CDB ops.\n\nAfter the recent se_task removal patches from hch, this value for IBLOCK\nbackends being set via configfs by userspace from an saved max_sectors\nvalue that is turning out to be problematic, so it makes sense to go ahead\nand remove this now legacy attribute all-together.\n\nThis patch also continues to make se_dev_set_default_attribs() do\n(sectors / block_size) alignment for what actually get used by\ntarget_core_mod to be safe here, following the same alignment currently\nused by fabric_max_sectors.\n\nReported-by: Andy Grover \u003cagrover@redhat.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "d5dc28eb92f2a2305a02cb3a5f1ed36542d47512", "tree": "97f2721b2eaa270a329d4dd556cd254a94bdb41b", "parents": [ "d43d6aea844e66b847e3a0e5cb2c24b6ba84908a" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Apr 24 00:25:09 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:11:28 2012 -0700" }, "message": "target: remove the t_se_count field in struct se_cmd\n\nNow that tasks are gone we are guaranteed to only get a single completion\nper command, and thus don\u0027t need this counter.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "d43d6aea844e66b847e3a0e5cb2c24b6ba84908a", "tree": "8a9a365b75d1bd186f29287a09702ba4e27b0c33", "parents": [ "226b6faf2e30e272ee5a52c2ed0a8c8a1d7b47d1" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Apr 24 00:25:08 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:11:27 2012 -0700" }, "message": "target: remove the t_task_cdbs_ex_left field in struct se_cmd\n\nNow that tasks are gone we are guaranteed to only get a single completion\nper command, and thus don\u0027t need this counter.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "226b6faf2e30e272ee5a52c2ed0a8c8a1d7b47d1", "tree": "f39c5904c6da5f04a6f5860c646005e21fba4282", "parents": [ "5787cacd0bd5ee016ad807b244550d34fe2beebe" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Apr 24 00:25:07 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:11:27 2012 -0700" }, "message": "target: remove the t_task_cdbs_left field in struct se_cmd\n\nNow that tasks are gone we are guaranteed to only get a single completion\nper command, and thus don\u0027t need this counter.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "5787cacd0bd5ee016ad807b244550d34fe2beebe", "tree": "d9303fd0c5767f945c13dfc34eb0b16a4f4d2730", "parents": [ "cf572a9627c9ae86082216de109780c1d2e2ee28" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Apr 24 00:25:06 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:11:26 2012 -0700" }, "message": "target: remove struct se_task\n\nWe can use struct se_cmd for everything it did. Make sure to pass the S/G\nlist and data direction to the execution function to ease adding back BIDI\nsupport later on.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "cf572a9627c9ae86082216de109780c1d2e2ee28", "tree": "fe6eec75bc3c388534390efe3be2b3b2c4c93903", "parents": [ "785fdf70b2b46588c973ad8b65ed62871994452f" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Apr 24 00:25:05 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:11:26 2012 -0700" }, "message": "target: move the state and execute lists to the command\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "785fdf70b2b46588c973ad8b65ed62871994452f", "tree": "e564a42e34e5a471f6d69a3d8e5e845ebf8c8eef", "parents": [ "4101f0a89d4eb13f04cb0344d59a335b862ca5f9" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Apr 24 00:25:04 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:11:25 2012 -0700" }, "message": "target: simplify command to task linkage\n\nNow that we only have a single task per command we can use a direct pointer\nto it instead of list.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "6bb35e009b656b36f7985057822c5fbf53ea75b7", "tree": "31384150dbd2828ee72a0802efb23692511f032f", "parents": [ "bd31377878868a47b2d2e6f570fb0cace828555b" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Apr 23 11:35:33 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:11:14 2012 -0700" }, "message": "target: replace -\u003eexecute_task with -\u003eexecute_cmd\n\nMake CDB emulation work on commands instead of tasks again as a preparation\nof removing tasks completely.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "bd31377878868a47b2d2e6f570fb0cace828555b", "tree": "f42aca1e190ab4bf4a1cca98e5a354b81a9a33c2", "parents": [ "7a83aa4e7b26ba61b9dc98d5c0bf725f4df0043f" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Apr 23 11:35:32 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:03:46 2012 -0700" }, "message": "target: remove the task_sectors field in struct se_task\n\nRemove the task_sectors field that isn\u0027t used anywhere.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "7a83aa4e7b26ba61b9dc98d5c0bf725f4df0043f", "tree": "dc2c7e29f7850e238e5f752aaeb4bbaa8dd69f94", "parents": [ "72a0e5e2e2e6b577343c79013611ad5701e94482" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Apr 23 11:35:31 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:03:41 2012 -0700" }, "message": "target: remove the task_size field in struct se_task\n\nNow that we don\u0027t split commands the size field in the task is always\nequivalent to the one in the CDB, even in cases where we have two tasks\ndue to a BIDI transfer. Just refer the the size in the command instead\nof duplicating it in the task.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "72a0e5e2e2e6b577343c79013611ad5701e94482", "tree": "06e41f5240a98040a2a6eed30ab75c0c4213baf2", "parents": [ "ed3102c678c6274933948f9097bb4e2b314e5f82" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Apr 23 11:35:30 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:02:59 2012 -0700" }, "message": "target: remove the task_lba field in struct se_task\n\nNow that we don\u0027t split commands the lba field in the task is always\nequivalent to the one in the CDB, even in cases where we have two tasks\ndue to a BIDI transfer. Just refer the the lba in the command instead\nof duplicating it in the task.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "ed3102c678c6274933948f9097bb4e2b314e5f82", "tree": "a1dabb18032a61539905fe932bc1949cb6e8dd80", "parents": [ "32d8d2e4300cabb0acbf4b60fa695924be27d0d1" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Apr 23 11:35:29 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun May 06 15:02:59 2012 -0700" }, "message": "target: remove target_get_task_cdb\n\nNow that tasks are always the same size as the command there is no need\nto rewrite a CDB in common code. Notw that we keep the separately allocated\nCDB in the pscsi and stgt backends for now, to easy reintroducing any\ncommand splitting local to these backends if nessecary.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "a12f41f8412ff57057906ebbe146fda37db158ac", "tree": "e3138df9188c5fff3156e53c8f6781240ee062cb", "parents": [ "8b1e1244db85d58f7c612870ec2c1afd9098ae93" ], "author": { "name": "Andy Grover", "email": "agrover@redhat.com", "time": "Tue Apr 03 15:51:20 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Apr 14 17:40:36 2012 -0700" }, "message": "target: Rename target_allocate_tasks to target_setup_cmd_from_cdb\n\nThis patch renames a horribly misnamed function that no longer allocate\ntasks to something more descriptive for it\u0027s modern use in target core.\n\n(nab: Fix up ib_srpt to use this as well ahead of a target_submit_cmd\nconversion)\n\nSigned-off-by: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "957525a2424aad367d6e0efb64e440b2b37fa5cd", "tree": "dc56eaa7a6e60ad674059d110e6d4145adb55efd", "parents": [ "6f9e7f01b69bc2d6021e4ab827e1fd78c82cbc52" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Fri Mar 30 11:29:15 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Apr 14 17:40:32 2012 -0700" }, "message": "target: Remove transport_do_task_sg_chain() and associated detritus\n\nNow that all fabrics are converted over to using se_cmd-\u003et_data_sg\ndirectly, we can drop the task sg chaining support. With the modern\nmemory allocation in target core, task sg chaining is needless\noverhead -- we would split up the main cmd sglist into pieces, and\nthen splice those pieces back together instead of just using the\noriginal list directly.\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "b0d7994660af1601cc26ef7ab748569fdb9c253b", "tree": "49740bdc2f2284762b1d367295b238c5f3bc931d", "parents": [ "58d926187758f76647d68e2f51ce44e6ace81873" ], "author": { "name": "Sebastian Andrzej Siewior", "email": "bigeasy@linutronix.de", "time": "Tue Jan 10 14:16:59 2012 +0100" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Apr 14 17:40:03 2012 -0700" }, "message": "target: add unknown size flag to target_submit_cmd()\n\nThe UASP protocol does not inform the target device upfront how much\ndata it should expect so we have to learn in from the CDB. So in order\nto handle this case, add a TARGET_SCF_UNKNOWN_SIZE to target_submit_cmd()\nand perform an explictly assignment for se_cmd-\u003edata_length from the\nextracted CDB size in transport_generic_cmd_sequencer().\n\nSigned-off-by: Sebastian Andrzej Siewior \u003cbigeasy@linutronix.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "58d926187758f76647d68e2f51ce44e6ace81873", "tree": "c877294c394ab14cbdeca4af1a3e7835d644f73b", "parents": [ "0034102808e0dbbf3a2394b82b1bb40b5778de9e" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Mar 20 21:26:48 2012 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Apr 14 16:04:58 2012 -0700" }, "message": "target: Remove obsolete DF_READ_ONLY usage\n\nThis was used at one time as a hack by FILEIO backend registration to\nallow a struct block_device that was claimed with blkdev_get (by a local\nfilesystem mount for example) to be exported as read-only (SCSI WP\u003d1).\n\nSince FILEIO backend registration will no longer attempt to obtain\nexclusive access to an underlying struct block_device here, this flag is\nnow obsolete.\n\nReported-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "f2083241f23722207676025abbb45a301d412e69", "tree": "c340376faf2a1eebe0983075adee2a3c8070c064", "parents": [ "4a5a75f32dcbcd0b2685f74fd4ede26edf8765a9" ], "author": { "name": "Jörn Engel", "email": "joern@logfs.org", "time": "Thu Mar 15 15:05:40 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Thu Mar 15 19:15:51 2012 -0700" }, "message": "target: Use array_zalloc for device_list\n\nTurns an order-8 allocation into slab-sized ones, thereby preventing\nallocation failures with memory fragmentation.\n\nThis likely saves memory as well, as the slab allocator can pack objects\nmore tightly than the buddy allocator.\n\n(nab: Fix lio-core patch fuzz)\n\nSigned-off-by: Joern Engel \u003cjoern@logfs.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "4a5a75f32dcbcd0b2685f74fd4ede26edf8765a9", "tree": "9c05e0df5d2cfe06ed490ce763347f65aaf64c6a", "parents": [ "b168fe8cfe530daabbdf632af4554600006a81a9" ], "author": { "name": "Jörn Engel", "email": "joern@logfs.org", "time": "Thu Mar 15 15:05:12 2012 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Thu Mar 15 19:15:38 2012 -0700" }, "message": "target: Use array_zalloc for tpg_lun_list\n\nTurns an order-10 allocation into slab-sized ones, thereby preventing\nallocation failures with memory fragmentation.\n\nThis likely saves memory as well, as the slab allocator can pack objects\nmore tightly than the buddy allocator.\n\nSigned-off-by: Joern Engel \u003cjoern@logfs.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "9765b1f327951e643a97f43924d1fbeb4e90ffba", "tree": "d1e9691a7b366874ca37bf229f84443384fd5366", "parents": [ "031ed4d565b31880a4136bb7366bc89f5b1dba7d" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:50:19 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:55:17 2012 -0800" }, "message": "target: Bump core version to v4.1.0-rc2-ml + fabric versions\n\nBump core version to v4.1.0-rc2-ml, and for versions from the\nfollowing mainline fabric modules:\n\nloopback: v2.1-rc2\ntcm_fc: v0.4\niscsi-target: v4.1.0-rc2\n\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "c7ec05c82bfd6acf1fd800d628591500805f3179", "tree": "c64d5d330120d038f5ee9fe90b7a7eba0a0b3459", "parents": [ "99367f01b8df2018ff3328ddf4111e701d34188d" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Feb 28 00:22:12 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:42:55 2012 -0800" }, "message": "target: Drop unused legacy target_core_fabric_ops API callers\n\nThis patch drops the following unused legacy API callers from target_core_fabric.h:\n\n*) TFO-\u003efall_back_to_erl0()\n*) TFO-\u003estop_session()\n*) TFO-\u003esess_logged_in()\n*) TFO-\u003eis_state_remove()\n\nThis patch also removes the stub usage in loopback, tcm_fc, iscsi_target,\nand ib_srpt fabric modules.\n\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "337c060701fb3d27d9945bf7af7ba194ae2153a8", "tree": "7b2d50a7cc4b291a876e95c60388cf57a52d912a", "parents": [ "afb999ffc48f5e7ec18e6f8f9e68aa3d0085862d" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:36:21 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:42:19 2012 -0800" }, "message": "target: Convert se_node_acl-\u003eacl_group removal to use -\u003eacl_kref\n\nThis patch converts core_tpg_del_initiator_node_acl() shutdown from configfs\ncontext to use se_node_acl-\u003eacl_kref and -\u003eacl_free_comp in order to wait for\noutstanding fabric callbacks to complete via transport_deregister_session()\ncallbacks before waking -\u003eacl_free_comp from the last -\u003eacl_kref put.\n\nIt also changes core_tpg_del_initiator_node_acl() to setup a local sess_list\nwith target_get_session() + acl-\u003eacl_stop \u003d 1 for active sessions that will\nbe shutdown, and changes transport_deregister_session_configfs() to check\nfor -\u003eacl_stop usage.\n\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Joern Engel \u003cjoern@logfs.org\u003e\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "afb999ffc48f5e7ec18e6f8f9e68aa3d0085862d", "tree": "781a1055eaed8bef8148c95fa6f9e779484511c6", "parents": [ "01468346546a9bcc09035a7e8b71f78af5b7133a" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Thu Mar 08 23:45:02 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:42:09 2012 -0800" }, "message": "target: Add se_node_acl-\u003eacl_kref for -\u003eacl_free_comp usage\n\nThis patch adds se_node_acl-\u003eacl_kref for use with -\u003eacl_free_comp\nduring explict se_node_acl release. It adds kref_init() during\nse_node_acl setup, kref_get() during __transport_register_session()\n-\u003e target_put_nacl() with existing transport_deregister_session()\nfabric callback usage.\n\nIt also moves transport_free_session() to release *se_sess memory\nafter target_put_nacl() execution in transport_deregister_session()\n\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Joern Engel \u003cjoern@logfs.org\u003e\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "01468346546a9bcc09035a7e8b71f78af5b7133a", "tree": "5d7b69d10a2fb7e7f34717531fd099cae24f8044", "parents": [ "41ac82b668177876d5965d5f60956fb0fbcdb514" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:32:52 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:41:59 2012 -0800" }, "message": "target: Add se_node_acl-\u003eacl_free_comp for NodeACL release path\n\nAdd se_node_acl-\u003eacl_free_comp for NodeACL release path to wait for outstanding\nfabric session shutdown to complete in transport_deregister_session() before\nfinishing NodeACL release from configfs process context.\n\nAlso make transport_deregister_session() clear the comp_nacl bit\nto skip se_node_acl-\u003eacl_free_comp completion for dynamically generated\nNodeACL during fabric session shutdown.\n\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Joern Engel \u003cjoern@logfs.org\u003e\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "41ac82b668177876d5965d5f60956fb0fbcdb514", "tree": "229c299330fa9c3cb4cb5a1903e0559e3e7a5d5b", "parents": [ "140854cb72525246745b67300d35101ad2875a39" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun Feb 26 22:22:10 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Mar 10 14:41:53 2012 -0800" }, "message": "target: Add se_sess-\u003esess_kref + get/put helpers\n\nThis patch adds basic se_session-\u003esess_kref and get/put helpers for fabric\nsession reference counting. It sets the initial kref in transport_init_session()\nand adds a target_release_session() callback to invoke TFO-\u003eclose_session()\nfor final session shutdown.\n\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Joern Engel \u003cjoern@logfs.org\u003e\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "c0974f89385970455a60a999ae4fc0a783cf458d", "tree": "14abc856524251bf392bcaf0b2b8c621500ec1b2", "parents": [ "9f0d05c2c6185d7b5ba08a6c052e90837101031f" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 05:10:04 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 18:39:13 2012 -0800" }, "message": "target: Allow target_submit_tmr interrupt context + pass ABORT_TASK tag\n\nThis patch allows target_submit_tmr() to pass gfp_t for se_cmd-\u003ese_tmr_req\nallocation, and also set up se_cmd-\u003ese_tmr_req-\u003eref_task_tag for passed\ntag with TMR_ABORT_TASK.\n\nAlso update tcm_fc(fcoe) parameter usgae and add ref_task_tag FIXME\nfor TMR_ABORT_TASK usage,\n\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nCc: Kiran Patil \u003ckiran.patil@intel.com\u003e\nCc: Arun Easi \u003carun.easi@qlogic.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "c7042cae58c13970f39c0820a3aab2a13dda9fe1", "tree": "63d75c4c8e37d9a26033ed11b47345b4ae697524", "parents": [ "3de55ec74aa803d7aa1e581700086d755c2599d2" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 01:40:24 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 18:38:34 2012 -0800" }, "message": "target: Fix target_submit_tmr se_tmr_req allocation failures\n\nThis patch makes target_submit_tmr() se_tmr_req allocation occur before\ntarget_get_sess_cmd(), and changes target_submit_tmr() to return a failure\nw/ non zero status to the fabric caller upon core_tmr_alloc_req() failure.\n\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nCc: Kiran Patil \u003ckiran.patil@intel.com\u003e\nCc: Arun Easi \u003carun.easi@qlogic.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "015487b89f27d91d95a056cdc3c85e6c729bff12", "tree": "0cecb2acc903154e25abb23e8f345f301fcd1ef5", "parents": [ "effc6cc8828257c32c37635e737f14fd6e19ecd7" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Feb 13 16:18:17 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:49 2012 -0800" }, "message": "target: Untangle front-end and back-end meanings of max_sectors attribute\n\nse_dev_attrib.max_sectors currently has two independent meanings:\n\n - It is reported in the block limits VPD page as the maximum transfer\n length, ie the largest IO that the front-end (fabric) can handle.\n Also the target core doesn\u0027t enforce this maximum transfer length.\n\n - It is used to hold the size of the largest IO that the back-end can\n handle, so we know when to split SCSI commands into multiple tasks.\n\nFix this by adding a new se_dev_attrib.fabric_max_sectors to hold the\nmaximum transfer length, and checking incoming IOs against that limit.\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "d95b82461c56a6ff8ff248b101049a69ebb20278", "tree": "0a0c8ad9d37c529c859ddcdfc7123b80bd195b0c", "parents": [ "3d28934aaae5e924afedf0f5cb42e1316514da6b" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Feb 13 16:18:14 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:49 2012 -0800" }, "message": "target: Fix up handling of short INQUIRY buffers\n\nIf the initiator sends us an INQUIRY command with an allocation length\nthat\u0027s shorter than what we want to return, we\u0027re simply supposed to\ntruncate our response and return what the initiator gave us space for,\nwithout signaling any error. Current target code has various tests that\ndon\u0027t fill out the full response if the buffer is too short and\nsometimes return errors incorrectly.\n\nFix this up by allocating a bounce buffer for INQUIRY responses if we\nneed to, ie if we have cmd-\u003edata_length too small as well as\nSCF_PASSTHROUGH_SG_TO_MEM_NOALLOC set in cmd-\u003ese_cmd_flags -- for most\nfabrics, we always allocate at least a full page, but for tcm_loop we\nmay have a small buffer coming directly from the SCSI stack.\n\nThis lets us delete a lot of cmd-\u003edata_length checking, and also makes\nour INQUIRY handling correct per SPC in a lot more cases.\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "86715569d085addc635c2b55ee8acb79d3a7fbbf", "tree": "b6b6d554b329a86845d8de661529c3fc2bc6dbef", "parents": [ "2fbff1276bb635294b3384b6c167f72d0acdab95" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Feb 13 01:07:22 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:49 2012 -0800" }, "message": "target: Add SCF_ACK_KREF flag for acknowledgement kref\n\nWhen TARGET_SCF_ACK_KREF is in use with target_submit_cmd() for\nsetting the extra acknowledgement reference to se_cmd-\u003ecmd_kref,\ngo ahead and set SCF_ACK_KREF in order to be used later by\nabort task.\n\nReported-by: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "2fbff1276bb635294b3384b6c167f72d0acdab95", "tree": "d480bcaff6665583ff5f8120cc81754a8f3651b0", "parents": [ "a1edf9cf6d1efed2981fcf6e4caa704fbb1bd093" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Feb 10 16:18:11 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:48 2012 -0800" }, "message": "target: Export transport_generic_request_failure symbol\n\ntransport_generic_request_failure() is a wrapper around calling\ntransport_send_check_condition_and_sense() that is required once\nan se_cmd-\u003ecmd_kref has been obtained via target_submit_cmd() -\u003e\ntarget_get_sess_cmd().\n\ntcm_qla2xxx currently requires this, and since it\u0027s necessary for\nother callers using target_submit_cmd() make it exportable now.\n\nReported-by: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "ea98d7f9c7cc38de55f81a1c249112442b93c7c9", "tree": "cd5b136fa5b22ba290a9da8992c486a5d5ff68e8", "parents": [ "a1321f71e87930579afc4a4029cce128c23f3fd3" ], "author": { "name": "Andy Grover", "email": "agrover@redhat.com", "time": "Thu Jan 19 13:39:21 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:48 2012 -0800" }, "message": "target: Add target_submit_tmr helper function\n\nSimilar to target_submit_cmd, this function lets fabrics call one function\n(albeit with a lot of parameters) instead of 3 or more.\n\n(nab: Add missing return for transport_lookup_tmr_lun failure)\n\nSigned-off-by: Andy Grover \u003cagrover@redhat.com\u003e\nCc: Kiran Patil \u003ckiran.patil@intel.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "c8e31f26feeb03dc6f51bff68135cc58431e099b", "tree": "8d584fcb7cb7a58988cd2463a9a26f577410e6ac", "parents": [ "35b2cdc4fea1f0d13e1602c07e62c797c9fe5ed4" ], "author": { "name": "Andy Grover", "email": "agrover@redhat.com", "time": "Thu Jan 19 13:39:17 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:47 2012 -0800" }, "message": "target: Add SCF_SCSI_TMR_CDB usage and drop se_tmr_req_cache\n\nChange the test for if a cmd is a tmr request to checking if\nSCF_SCSI_TMR_CDB (a new flag) is set in cmd-\u003ese_cmd_flags.\n\nAlso remove se_tmr_req_cache usage in favor of kzalloc usage,\nand make core_tmr_alloc_req() return int + setup se_cmd-\u003ese_tmr_req\ndirectly and fix up various fabric module usages\n\nCc: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "ef28640497f5935bf614e1e29b16972d8b97cdb2", "tree": "76ef67c94005581784fcd943e293c5ea4f403c75", "parents": [ "644df699d598c12989238b3f14fdabf78c6076ab" ], "author": { "name": "Andy Grover", "email": "agrover@redhat.com", "time": "Thu Jan 19 13:39:13 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:47 2012 -0800" }, "message": "target: Remove unused struct se_queue_req\n\nSigned-off-by: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "644df699d598c12989238b3f14fdabf78c6076ab", "tree": "ffe92b9148bc3209a183b668e6786ca6a28b187c", "parents": [ "e35fa8c2d0feb977c2f7d14a973b4132483ffef3" ], "author": { "name": "Andy Grover", "email": "agrover@redhat.com", "time": "Thu Jan 19 13:39:12 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:46 2012 -0800" }, "message": "target: fix comment typos\n\nSigned-off-by: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "3d26fea01d5f80e3e585d69d8d73a60e1ca563a0", "tree": "6955955e14533854f9e9eaae7b7475e819bc53d5", "parents": [ "7d680f3b74dd6f0f57569eeeee8c257790ceaa96" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Wed Dec 21 14:14:05 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:46 2012 -0800" }, "message": "target: remove the transport_lun_active field in struct se_cmd\n\nThere is no reason to have a flag telling if a command is on the per-lun list,\nwe can simply do a list_empty check before removing it as long as we\u0027re careful\nto always use list_del_init.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "7d680f3b74dd6f0f57569eeeee8c257790ceaa96", "tree": "e0797ab722c222213b7cbe4fa9ba4a61db2d8b0c", "parents": [ "b01543dfe67bb1d191998e90d20534dc354de059" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Wed Dec 21 14:13:47 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Feb 25 14:37:45 2012 -0800" }, "message": "target: replace various cmd flags with a transport state\n\nReplace various atomic_ts used as flags in struct se_cmd with a single\ntransport_state bitmap that requires t_state_lock to be held for modifications.\n\nIn the target core that assumption generally is true, but some recently added\ncode in the SRP target had to grow new lock calls. I can\u0027t say I like the way\nhow it messes with the command state directly, but let\u0027s leave that for later.\n\n(Re-add missing ib_srpt.c changes that nab dropped..)\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "1edcdb497ef418122cd4f98e157660cf594b345a", "tree": "bf41780935b8996f85a0ae090897495961deb6d7", "parents": [ "95fe1ee41e23fa271416da67483594dde74bc6ca" ], "author": { "name": "Andy Grover", "email": "agrover@redhat.com", "time": "Thu Jan 19 13:39:23 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Feb 07 06:41:04 2012 +0000" }, "message": "target: Change target_submit_cmd() to return void\n\nRetval not very useful, and may even be harmful. Once submitted, fabrics\nshould expect a sense error if anything goes wrong. All fabrics checking\nof this retval are useless or broken:\n\nfc checks it just to emit more debug output.\nib_srpt trickles retval up, then it is ignored.\nqla2xxx trickles it up, which then causes a bug because the abort goto\nin qla_target.c thinks cmd hasn\u0027t been sent to target.\n\nJust returning nothing is best.\n\nSigned-off-by: Andy Grover \u003cagrover@redhat.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "4949314c7283ea4f9ade182ca599583b89f7edd6", "tree": "bbd2316996bc2f19c1b3a8c24f8591f12d2330fa", "parents": [ "e8904dc5008ef92f0f62391d6557f03f921eeb32" ], "author": { "name": "Andy Grover", "email": "agrover@redhat.com", "time": "Mon Jan 16 16:57:08 2012 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Jan 18 08:35:58 2012 +0000" }, "message": "target: Allow control CDBs with data \u003e 1 page\n\nWe need to handle \u003e1 page control cdbs, so extend the code to do a vmap\nif bigger than 1 page. It seems like kmap() is still preferable if just\na page, fewer TLB shootdowns(?), so keep using that when possible.\n\nRename function pair for their new scope.\n\nSigned-off-by: Andy Grover \u003cagrover@redhat.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "895f3022523361e9b383cf48f51feb1f7d5e7e53", "tree": "6a692ebaacc4af38c8869eee0da7c6e868b1232a", "parents": [ "6d5b59756033c2a029bde7262bb5f8d45f4ca952" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Dec 13 14:55:33 2011 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Dec 16 06:29:04 2011 +0000" }, "message": "target: Set additional sense length field in sense data\n\nThe target code was not setting the additional sense length field in the\nsense data it returned, which meant that at least the Linux stack\nignored the ASC/ASCQ fields. For example, without this patch, on a\ntcm_loop device:\n\n # sg_raw -v /dev/sda 2 0 0 0 0 0\n\ngives\n\n cdb to send: 02 00 00 00 00 00\n SCSI Status: Check Condition\n\n Sense Information:\n Fixed format, current; Sense key: Illegal Request\n Raw sense data (in hex):\n 70 00 05 00 00 00 00 00\n\nwhile after the patch we correctly get the following (which matches what\na regular disk returns):\n\n cdb to send: 02 00 00 00 00 00\n SCSI Status: Check Condition\n\n Sense Information:\n Fixed format, current; Sense key: Illegal Request\n Additional sense: Invalid command operation code\n Raw sense data (in hex):\n 70 00 05 00 00 00 00 0a 00 00 00 00 20 00 00 00\n 00 00\n\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\nCc: stable@kernel.org\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "4d2300ccffd22d1d0213b6a8e4d685eb6ca069c0", "tree": "a76817aa4aaba5a6ffb05b3e6a0ea9ca654a0519", "parents": [ "65586d51e0986be574118286c3d0007e903a2add" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Nov 30 18:18:33 2011 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:48:46 2011 +0000" }, "message": "target: Remove extra se_device-\u003eexecute_task_lock access in fast path\n\nThis patch makes __transport_execute_tasks() perform the addition of\ntasks to dev-\u003eexecute_task_list via __transport_add_tasks_from_cmd()\nwhile holding dev-\u003eexecute_task_lock during normal I/O fast path\nsubmission.\n\nIt effectively removes the unnecessary re-acquire of dev-\u003eexecute_task_lock\nduring transport_execute_tasks() -\u003e transport_add_tasks_from_cmd() ahead\nof calling __transport_execute_tasks() to queue tasks for the passed\n*se_cmd descriptor.\n\n(v2: Re-add goto check_depth usage for multi-task submission for now..)\n\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Joern Engel \u003cjoern@logfs.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "65586d51e0986be574118286c3d0007e903a2add", "tree": "0684f76d53e666ffe2a2c77f8e3947d263fa1b75", "parents": [ "40be67f4c588fe2f3e2dbd60ae1f470abc5b6ad8" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Nov 30 01:25:21 2011 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:42:13 2011 +0000" }, "message": "target: Drop se_device TCQ queue_depth usage from I/O path\n\nHistorically, pSCSI devices have been the ones that required target-core\nto enforce a per se_device-\u003edepth_left. This patch changes target-core\nto no longer (by default) enforce a per se_device-\u003edepth_left or sleep in\ntransport_tcq_window_closed() when we out of queue slots for all backend\nexport cases.\n\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Joern Engel \u003cjoern@logfs.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "ec54cc081ead14e85736c6543e36ec59be8a7d3d", "tree": "8164a0a18787602a89eaaf83c0861676821fdada", "parents": [ "4355a9110eeb2eaf1dd44fcab16ccbd1c8c5fad4" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 10:50:16 2011 +0000" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:42:11 2011 +0000" }, "message": "target: Remove TFO-\u003echeck_release_cmd() fabric API caller\n\nRemove the now unused target_core_fabric_ops-\u003echeck_release_cmd() as\ntarget_core handles this directly for se_cmd-\u003ecmd_kref objects now.\n\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "a63607855224702ea17e6016ecf3f7d544e83625", "tree": "844d9a7dbdc636ba657763f04945b91ca8da9ba7", "parents": [ "7481deb413be132a22193e8a0bce88b311ecb3c2" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Nov 18 20:36:22 2011 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:40:56 2011 +0000" }, "message": "target: Add target_submit_cmd() for process context fabric submission\n\nThis patch adds a target_submit_cmd() caller that can be used by fabrics\nto submit an uninitialized se_cmd descriptor to an struct se_session +\nunpacked_lun from workqueue process context. This call will invoke the\nfollowing steps:\n\n- transport_init_se_cmd() to setup se_cmd specific pointers\n- Obtain se_cmd-\u003ecmd_kref references with target_get_sess_cmd()\n- set se_cmd-\u003et_tasks_bidi\n- transport_lookup_cmd_lun() to setup struct se_cmd-\u003ese_lun from\n the passed unpacked_lun\n- transport_generic_allocate_tasks() to setup the passed *cdb, and\n- transport_handle_cdb_direct() handle READ dispatch or WRITE\n ready-to-transfer callback to fabric\n\nv2 changes from hch feedback:\n\n*) Add target_sc_flags_table for target_submit_cmd flags\n*) Rename bidi parameter to flags, add TARGET_SCF_BIDI_OP\n*) Convert checks to BUG_ON\n*) Add out_check_cond for transport_send_check_condition_and_sense\n usage\n\nv3 changes:\n\n*) Add TARGET_SCF_ACK_KREF for target_submit_cmd into\n target_get_sess_cmd to determine when the fabric caller is expecting\n a second kref_put() from fabric packet acknowledgement.\n\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "7481deb413be132a22193e8a0bce88b311ecb3c2", "tree": "6d417c6856f983cb28fae0717563f4efcd6019e1", "parents": [ "ce136176fea522fc8f4c16dcae7e8ed1d890ca39" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Nov 12 00:32:17 2011 -0800" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:38:29 2011 +0000" }, "message": "target: Make target_put_sess_cmd use target_release_cmd_kref\n\nThis patch moves target_put_sess_cmd() to use a se_cmd-\u003ecmd_kref\ncallback target_release_cmd_kref when performing driver release of\nfabric-\u003ese_cmd descriptor memory. It sets the default cmd_kref\ncount value to \u00272\u0027 within target_get_sess_cmd() setup, and\ncurrently assumes TFO-\u003echeck_stop_free() usage.\n\nIt drops se_tfo-\u003echeck_release_cmd() usage in the main\ntransport_release_cmd codepath.\n\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "e0a53e70e8d519579d92c150606ece22333645cb", "tree": "7f28469be2885fd7979ec584dcc65b3b35748871", "parents": [ "2e88efd3aaafa0df05593cdfa5fd17afe922781d" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Nov 29 03:29:38 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:28:13 2011 +0000" }, "message": "target: remove overagressive ____cacheline_aligned annoations\n\nIf we want dynamically allocated objects to be cacheline aligned we need\nto tell that to the slab allocator by using the proper flags and not\nby liberally sprinkling annotations onto all structures.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "1880807adb21d741f08b747956c90bf4a6f95fbf", "tree": "fb3815bf85679116bb455ded1a678049950a3d9f", "parents": [ "41e16e981679124c78c30f046d4f0b71d86ff1b2" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Wed Nov 23 06:54:36 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:27:02 2011 +0000" }, "message": "target: make the se_task task_state_active a normal bool\n\nThere is no need to make task_state_active an atomic_t given that it is\nalways set under the execute_task_lock so we can make it a simple bool.\nAlso rename it to t_state_active to be closer to the list it guards,\nand make sure all checks before the list addion/removal actually happen\nunder execute_task_lock.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "41e16e981679124c78c30f046d4f0b71d86ff1b2", "tree": "2db886814e0f8b9b39d8e8b836e3598d128d14e1", "parents": [ "ef804a849ffae058a783e9dddd24cc1a555bbdb4" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Wed Nov 23 06:54:15 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:26:44 2011 +0000" }, "message": "target: remove the se_task task_error_status field\n\nWe only reach transport_complete_task once per task, so the test and set on\ntask_error_status is never going to have an effect.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "ef804a849ffae058a783e9dddd24cc1a555bbdb4", "tree": "e67737300936c2b94359ded96d9113981cb610aa", "parents": [ "c4795fb20edf2fe2c862c8fe9f8b681edeb79ac1" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Wed Nov 23 06:53:58 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:26:27 2011 +0000" }, "message": "target: fold se_task.task_sense into task_flags\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "c4795fb20edf2fe2c862c8fe9f8b681edeb79ac1", "tree": "506a8cb0a12ea2b56c5ebed5e6bc2a0d842c069d", "parents": [ "e26d99aed42ec6cdc9540d19c77ac5d4dd2c5b00" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Wed Nov 16 09:46:48 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 11:26:05 2011 +0000" }, "message": "target: header reshuffle, part2\n\nThis reorganized the headers under include/target into:\n\n - target_core_base.h stays as is with all target-wide data stuctures and defines\n - target_core_backend.h contains the whole interface to I/O backends\n - target_core_fabric.h contains the whole interface to fabric modules\n\nExcept for those only the various configfs macro headers stay around.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "e26d99aed42ec6cdc9540d19c77ac5d4dd2c5b00", "tree": "9c0424f81877d6c5cfaf7746447e0c9decf8e82e", "parents": [ "dc47ce90c3a822cd7c9e9339fe4d5f61dcb26b50" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Nov 14 12:30:30 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Dec 14 08:51:12 2011 +0000" }, "message": "target: reshuffle headers\n\nCreate a new headers, drivers/target/target_core_internal.h that is supposed\nto hold all target_core-internal prototypes. Move all non-exported includes\nfrom include/target to it, and merge the smaller prototype-only includes\ninside drivers/target into it as well. Mark functions that were found to\nnot be called outside their implementation file static.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "6f21475576dde397cd2580262209d4080fbd5458", "tree": "2fa1831eaec0722b2e48278e447a3aa57b52a9fc", "parents": [ "9649fa1b8764f64c8cc4293e197e14cd46fe7205" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Nov 29 03:29:59 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:57 2011 +0000" }, "message": "target: remove the unused se_dev_list\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "5c73b678f729ea087ef57b59a5d7b5dd3a97042b", "tree": "1dfbf0564bba408d6728cb2cf1054b3f0f2eddbc", "parents": [ "1289a0571c037b4757f60597d646aedb70361ec3" ], "author": { "name": "Jörn Engel", "email": "joern@logfs.org", "time": "Thu Nov 24 02:04:16 2011 +0100" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:56 2011 +0000" }, "message": "target: remove unused struct fields\n\nSome are never used, some are set but never read, dev_hoq_count is\nincremented and decremented, but never read.\n\nSigned-off-by: Joern Engel \u003cjoern@logfs.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "6fd126ffebef3897d8fca98644a9fd1cc5c7a5e3", "tree": "14371a3621bd18ed18ee5262844d673e4d48d843", "parents": [ "33c3fafc43e56a22be60ebe67bec5ba763d51010" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Nov 14 11:36:31 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:53 2011 +0000" }, "message": "target: remove the unused t_task_pt_sgl and t_task_pt_sgl_num se_cmd fields\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "33c3fafc43e56a22be60ebe67bec5ba763d51010", "tree": "87a482353c77ff469b4d7b4fab8df105087d8cf9", "parents": [ "2d3a4b51df4db2ee0415f42a63b9629a7977b975" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Nov 14 11:36:30 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:53 2011 +0000" }, "message": "target: remove the t_tasks_bidi se_cmd field\n\nAnd use a SCF_BIDI flag instead.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "2d3a4b51df4db2ee0415f42a63b9629a7977b975", "tree": "fd83d9881b4d2041061d62271fea53af8a81be44", "parents": [ "aad13ca20d960ab74b739d7bbe876dac4502f546" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Nov 14 11:36:29 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:52 2011 +0000" }, "message": "target: remove the t_tasks_fua se_cmd field\n\nAnd use a SCF_FUA flag instead.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "aad13ca20d960ab74b739d7bbe876dac4502f546", "tree": "be24d2405a4f53449a7368cbd8218edc06ad8828", "parents": [ "58a2801a4b9ad97d3685bb7a3344e17d60292908" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Nov 14 11:36:28 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:52 2011 +0000" }, "message": "target: remove the se_ordered_node se_cmd field\n\nWe never walk ordered_cmd_list in the se_device, so remove all code related\nto supporting it.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "58a2801a4b9ad97d3685bb7a3344e17d60292908", "tree": "57326befbef1dfdf86b20b2237939c333e1cddbc", "parents": [ "6297b07cbc42eb6b53eb88fce70a4727ea230797" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Nov 14 11:36:27 2011 -0500" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:52 2011 +0000" }, "message": "target: remove the se_obj_ptr and se_orig_obj_ptr se_cmd fields\n\nWe already have a perfectly valid se_device pointer in the command, so\nremove the mostly useless duplicates.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "5f655e8d2a7cdc41943f929e86054051d7441ec5", "tree": "1386eb5a5e2585ed4aff57a04c15ca341524dc7f", "parents": [ "330694a50f5b80e983136237c10516810fb427a9" ], "author": { "name": "Bart Van Assche", "email": "bvanassche@acm.org", "time": "Tue Nov 08 20:46:29 2011 +0100" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:50 2011 +0000" }, "message": "target: Avoid compiler warnings about signed one-bit bitfields\n\nConvert to unsigned bit fields for active I/O shutdown fields.\n\nSigned-off-by: Bart Van Assche \u003cbvanassche@acm.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "03e98c9eb916f3f0868c1dc344dde2a60287ff72", "tree": "cfa47c1cec5b0a589b868c34310ee25f6a6a1939", "parents": [ "5611cc4572e889b62a7b4c72a413536bf6a9c416" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Nov 04 02:36:16 2011 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Tue Dec 06 06:00:49 2011 +0000" }, "message": "target: Address legacy PYX_TRANSPORT_* return code breakage\n\nThis patch removes legacy usage of PYX_TRANSPORT_* return codes in a number\nof locations and addresses cases where transport_generic_request_failure()\nwas returning the incorrect sense upon CHECK_CONDITION status after the\nv3.1 converson to use errno return codes.\n\nThis includes the conversion of transport_generic_request_failure() to\nprocess cmd-\u003escsi_sense_reason and handle extra TCM_RESERVATION_CONFLICT\nbefore calling transport_send_check_condition_and_sense() to queue up\nresponse status. It also drops PYX_TRANSPORT_OUT_OF_MEMORY_RESOURCES legacy\nusgae, and returns TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE w/ a response\nfor these cases.\n\ntransport_generic_allocate_tasks(), transport_generic_new_cmd(), backend\nSCF_SCSI_DATA_SG_IO_CDB -\u003edo_task(), and emulated -\u003eexecute_task() have\nall been updated to set se_cmd-\u003escsi_sense_reason and return errno codes\nuniversally upon failure. This includes cmd-\u003escsi_sense_reason assignment\nin target_core_alua.c, target_core_pr.c and target_core_cdb.c emulation code.\n\nFinally it updates fabric modules to remove the legacy usage, and for\nTFO-\u003enew_cmd_map() callers forwards return values outside of fabric code.\niscsi-target has also been updated to remove a handful of special cases\nrelated to the cleanup and signaling QUEUE_FULL handling w/ ft_write_pending()\n\n(v2: Drop extra SCF_SCSI_CDB_EXCEPTION check during failure from\n transport_generic_new_cmd, and re-add missing task-\u003etask_error_status\n assignment in transport_complete_task)\n\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: stable@kernel.org\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "6aad3738f6a79fd0ca480eaceefe064cc471f6eb", "tree": "08fb9ec4824bf3320af01f29fe84b75f814c0fa0", "parents": [ "02ebbbd481635fd3ce7018e5bb19c18c0f1e4561", "5bda90c8f20f0af93375721533f4081a40fa6f41" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Nov 06 19:00:42 2011 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Nov 06 19:00:42 2011 -0800" }, "message": "Merge branch \u0027for-next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending\n\n* \u0027for-next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending:\n target: use -\u003eexectute_task for all CDB emulation\n target: remove SCF_EMULATE_CDB_ASYNC\n target: refactor transport_emulate_control_cdb\n target: pass the se_task to the CDB emulation callback\n target: split core_scsi3_emulate_pr\n target: split core_scsi2_emulate_crh\n target: Add generic active I/O shutdown logic\n target: add back error handling in transport_complete_task\n target/pscsi: blk_make_request() returns an ERR_PTR()\n target: Remove core TRANSPORT_FREE_CMD_INTR usage\n target: Make TFO-\u003echeck_stop_free return free status\n iscsi-target: Fix non-immediate TMR handling\n iscsi-target: Add missing CMDSN_LOWER_THAN_EXP check in iscsit_handle_scsi_cmd\n target: Avoid double list_del for aborted se_tmr_req\n target: Minor cleanups to core_tmr_drain_tmr_list\n target: Fix wrong se_tmr being added to drain_tmr_list\n target: Fix incorrect se_cmd assignment in core_tmr_drain_tmr_list\n target: Check -ENOMEM to signal QUEUE_FULL from fabric callbacks\n tcm_loop: Add explict read buffer memset for SCF_SCSI_CONTROL_SG_IO_CDB\n target: Fix compile warning w/ missing module.h include\n" }, { "commit": "d29a5b6acc4b63d4e05ff554509df6fbeaf527cd", "tree": "edd94dc0420982dea35a811827757b3145689df0", "parents": [ "6ed5a557905f1c4e9ca5f8a6d607303a12d097e1" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Thu Nov 03 17:50:44 2011 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Nov 04 10:43:35 2011 +0000" }, "message": "target: remove SCF_EMULATE_CDB_ASYNC\n\nAll -\u003eexecute_task instances now need to complete the I/O explicitly,\nwhich can either happen synchronously or asynchronously.\n\nNote that a lot of the CDB emulations appear to return success even if\nsome lowlevel operations failed. Given that this is an existing issue\nthis patch doesn\u0027t change that fact.\n\n(nab: Adding missing switch breaks in PR-IN + PR_OUT)\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "e76a35d6c809bd1638e3b1b535bb780ac731c380", "tree": "a67bbbbc16cb4e612015da8b0f86fe981d1a5f77", "parents": [ "617c0e06c1b30b799d8b25f92eefdc1b098cb9f8" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Thu Nov 03 17:50:42 2011 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Nov 04 08:00:17 2011 +0000" }, "message": "target: pass the se_task to the CDB emulation callback\n\nWe want to be able to handle all CDBs through it and remove hacks like\nalways using the first task in a CDB in target_report_luns.\n\nAlso rename the callback to -\u003eexecute_task to better describe its use.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "a17f091d1a7c570804cfc2c77701634da88f8ecf", "tree": "50884063148c31cb5ee7b9a3f13bcc41f585cb92", "parents": [ "2235007c4d3245c0eca5e49497aafe5a111c00fb" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Nov 02 21:52:08 2011 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Fri Nov 04 07:50:26 2011 +0000" }, "message": "target: Add generic active I/O shutdown logic\n\nThis patch adds the initial pieces of generic active I/O shutdown logic.\nThis is intended to be a \u0027opt-in\u0027 feature for fabric modules that\nincludes the following functions to provide a mechinism for fabric\nmodules to track se_cmd via se_session-\u003esess_cmd_list:\n\n*) target_get_sess_cmd() - Add se_cmd to sess-\u003esess_cmd_list, called\n from fabric module incoming I/O path.\n*) target_put_sess_cmd() - Check for completion or drop se_cmd from\n -\u003esess_cmd_list\n*) target_splice_sess_cmd_list() - Splice active I/O list from\n -\u003esess_cmd_list to -\u003esess_wait_list, can called with HW fabric\n lock held.\n*) target_wait_for_sess_cmds() - Walk -\u003esess_wait_list waiting on\n individual -\u003ecmd_wait_comp. Optional transport_wait_for_tasks()\n call.\n\ntarget_splice_sess_cmd_list() is allowed to be called under HW fabric\nlock, and performs the splice into se_sess-\u003esess_wait_list and set\nse_cmd-\u003ecmd_wait_set. Then target_wait_for_sess_cmds() walks the list\nwaiting for individual target_put_sess_cmd() fabric callbacks to\ncomplete.\n\nIt also adds TFO-\u003echeck_release_cmd() to split the completion and memory\nrelease calls, where a fabric module uses target_put_sess_cmd() to check\nfor I/O completion during session shutdown. This is currently pushed out\ninto fabric modules as current fabric code may sleep here waiting for\nTFO-\u003echeck_stop_free() to complete in main response path, and because\ntarget_wait_for_sess_cmds() calling TFO-\u003erelease_cmd() to free fabric\ndescriptor memory directly.\n\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nSigned-off-by: Nicholas A. Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "3151d069e9e77043b0e791719bc65896cf24d9f0", "tree": "7913988d9a7efc0ac7ef6cd44ca892243adcb9bd", "parents": [ "88dd9e26d6d3e743f9c7e4562b94b2ad3c2994d3" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Nov 02 08:28:20 2011 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Nov 02 15:58:46 2011 +0000" }, "message": "target: Remove core TRANSPORT_FREE_CMD_INTR usage\n\nThis patch drops TRANSPORT_FREE_CMD_INTR usage from target core, which\nincludes the removal of transport_generic_free_cmd_intr() symbol,\nTRANSPORT_FREE_CMD_INTR usage in transport_processing_thread(), and\nspecial case LUN_RESET handling to skip TRANSPORT_FREE_CMD_INTR processing\nin core_tmr_drain_cmd_list(). We now expect that fabric modules will\nuse an internal workqueue to provide process context when releasing\nse_cmd descriptor resources via transport_generic_free_cmd().\n\nReported-by: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Roland Dreier \u003croland@purestorage.com\u003e\nCc: Madhuranath Iyengar \u003cmni@risingtidesystems.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@risingtidesystems.com\u003e\n" }, { "commit": "88dd9e26d6d3e743f9c7e4562b94b2ad3c2994d3", "tree": "452154b4137170de74d4445c5fcbbf62ce2f04fd", "parents": [ "5a4c8666c6d576f076a7c6824589cdbb984c0f84" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Nov 02 03:33:16 2011 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Wed Nov 02 15:58:30 2011 +0000" }, "message": "target: Make TFO-\u003echeck_stop_free return free status\n\nThis patch converts target_core_fabric_ops-\u003echeck_stop_free() usage in\ntransport_cmd_check_stop() and associated fabric module usage to\nreturn \u00271\u0027 when the passed se_cmd has been released directly within\n-\u003echeck_stop_free(), or return \u00270\u0027 when the passed se_cmd has not\nbeen released.\n\nThis addresses an issue where transport_cmd_finish_abort() -\u003e\ntransport_cmd_check_stop_to_fabric() was leaking descriptors during\nLUN_RESET for modules using -\u003echeck_stop_free(), but not directly\nreleasing se_cmd in all cases.\n\nCc: stable@kernel.org\nSigned-off-by: Nicholas Bellinger \u003cnab@risingtidesystems.com\u003e\n" }, { "commit": "59e52534172d845ebffb0d7e85fc56fb7b857051", "tree": "49552e03f1bdb413cd8b5f7542e91770688d7047", "parents": [ "73692d9bb58ecc2fa73f4b2bfcf6eadaa6d49a26", "0d89e54c8249645404283436d952afc261a04e1e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Oct 25 12:11:02 2011 +0200" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Oct 25 12:11:02 2011 +0200" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (59 commits)\n MAINTAINERS: linux-m32r is moderated for non-subscribers\n linux@lists.openrisc.net is moderated for non-subscribers\n Drop default from \"DM365 codec select\" choice\n parisc: Kconfig: cleanup Kernel page size default\n Kconfig: remove redundant CONFIG_ prefix on two symbols\n cris: remove arch/cris/arch-v32/lib/nand_init.S\n microblaze: add missing CONFIG_ prefixes\n h8300: drop puzzling Kconfig dependencies\n MAINTAINERS: microblaze-uclinux@itee.uq.edu.au is moderated for non-subscribers\n tty: drop superfluous dependency in Kconfig\n ARM: mxc: fix Kconfig typo \u0027i.MX51\u0027\n Fix file references in Kconfig files\n aic7xxx: fix Kconfig references to READMEs\n Fix file references in drivers/ide/\n thinkpad_acpi: Fix printk typo \u0027bluestooth\u0027\n bcmring: drop commented out line in Kconfig\n btmrvl_sdio: fix typo \u0027btmrvl_sdio_sd6888\u0027\n doc: raw1394: Trivial typo fix\n CIFS: Don\u0027t free volume_info-\u003eUNC until we are entirely done with it.\n treewide: Correct spelling of successfully in comments\n ...\n" }, { "commit": "2e982ab92dff057c639d4a43ccfa275be62f5e59", "tree": "8c8467e06736efa042006b11210281894c75bd95", "parents": [ "415a090ade7e674018e3fa4255938e4c312339b3" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun Oct 23 18:46:36 2011 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Oct 24 03:22:08 2011 +0000" }, "message": "target: Remove legacy se_task-\u003etask_timer and associated logic\n\nThis patch removes the legacy usage of se_task-\u003etask_timer and associated\ninfrastructure that originally was used as a way to help manage buggy backend\nSCSI LLDs that in certain cases would never return back an outstanding task.\n\nThis includes the removal of target_complete_timeout_work(), timeout logic\nfrom transport_complete_task(), transport_task_timeout_handler(),\ntransport_start_task_timer(), the per device task_timeout configfs attribute,\nand all task_timeout associated structure members and defines in\ntarget_core_base.h\n\nThis is being removed in preparation to make transport_complete_task() run\nin lock-less mode.\n\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "415a090ade7e674018e3fa4255938e4c312339b3", "tree": "075991035ca3410cdd2a9110d2cde1c1ca485e4d", "parents": [ "af3f00c75949369d937f499f49118e879939724d" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sun Oct 23 18:16:13 2011 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Oct 24 03:22:06 2011 +0000" }, "message": "target: Fix incorrect transport_sent usage\n\nThis patch converts target-core to use se_cmd-\u003et_transport_sent instead of\na duplicated se_cmd-\u003etransport_sent member in a handful of locations.\nIt also updates iscsi_target to properly use -\u003et_transport_sent instead of\nit\u0027s own iscsi_cmd_t-\u003etransport_sent value that was not being assigned.\n\nReported-by: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "7c1c6af37af69a4ac4a6485c968496d257245b5d", "tree": "7837a4ec24de8dc4c373810137250c7e75d9f24b", "parents": [ "dbc5623eb2898f5b5dcdc0b16077bb3f58629c78" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Tue Oct 18 06:57:00 2011 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Oct 24 03:21:56 2011 +0000" }, "message": "target: remove the task_sg_bidi field se_task and pSCSI BIDI support\n\nThis field is never used given that BIDI handling happens at the\ncommand and not the task level. Remove it and the dead code in\npscsi that tries to work on it.\n\nIt also prevents pSCSI passthrough for the two currently enabled BIDI\ncommands now that task-\u003etask_sg_bidi support has been removed.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Boaz Harrosh \u003cbharrosh@panasas.com\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "dbc5623eb2898f5b5dcdc0b16077bb3f58629c78", "tree": "f69f4cba91f373a40bcfbbfca8b5383ee297320e", "parents": [ "35e0e757537b9239172e35db773dd062727fd612" ], "author": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Sat Oct 22 01:03:54 2011 -0700" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Oct 24 03:21:54 2011 +0000" }, "message": "target: transport_subsystem_check_init cleanups\n\nRemove the now unnecessary extra call to transport_subsystem_check_init() in\ntarget_core_register_fabric(), and also merge transport_subsystem_reqmods()\ndirectly into transport_subsystem_check_init().\n\nReported-by: Christoph Hellwig \u003chch@infradead.org\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "35e0e757537b9239172e35db773dd062727fd612", "tree": "3a54e1e426f1a962219325eac474b715a4699924", "parents": [ "59aaad1ec44d9a77c32b873b001f31c5af47fcc7" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Oct 17 13:56:53 2011 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Oct 24 03:21:52 2011 +0000" }, "message": "target: use a workqueue for I/O completions\n\nInstead of abusing the target processing thread for offloading I/O\ncompletion in the backends to user context add a new workqueue. This means\ncompletions can be processed as fast as available CPU time allows it,\nincluding in parallel with other completions and more importantly I/O\nsubmission or QUEUE FULL retries. This should give much better performance\nespecially on loaded systems.\n\nAs a fallout we can merge all the completed states into a single\none.\n\nOn the downside this change complicates lun reset handling a bit by\nrequiring us to cancel a work item only for those states that have it\ninitialized. The alternative would be to either always initialize the work\nitem to a dummy handler, or always use the same handler and do a switch on\nthe state. The long term solution will be a flag that says that the command\nhas an initialized work item, but that\u0027s only going to be useful once we\nhave more users.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" }, { "commit": "59aaad1ec44d9a77c32b873b001f31c5af47fcc7", "tree": "7246282fed654e4e09f79e4989e0b0689dd8e13b", "parents": [ "f2da9dbdb54f2e9fa00dd01af6ff2ab06b4d90b7" ], "author": { "name": "Christoph Hellwig", "email": "hch@infradead.org", "time": "Mon Oct 17 13:56:52 2011 -0400" }, "committer": { "name": "Nicholas Bellinger", "email": "nab@linux-iscsi.org", "time": "Mon Oct 24 03:21:51 2011 +0000" }, "message": "target: remove unused TRANSPORT_ states\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Nicholas Bellinger \u003cnab@linux-iscsi.org\u003e\n" } ], "next": "f2da9dbdb54f2e9fa00dd01af6ff2ab06b4d90b7" }