)]}' { "log": [ { "commit": "9993e0fe0f5f29c69e79efcb271ffc9843002985", "tree": "ed35a971df3613385819d58bf79817b7100c2992", "parents": [ "af0cdf4947818becfe209610b209315578645ab4" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Dec 07 20:32:40 2010 +0100" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Dec 12 15:47:02 2010 +0100" }, "message": "firewire: ohci: fix regression with Agere FW643 rev 06, disable MSI\n\nAgere FW643 rev 06, listed as \"11c1:5901 (rev 06) (prog-if 10 [OHCI])\",\nproduced SBP-2 I/O errors since kernel 2.6.36. Disabling MSI fixes it.\n\nSince MSI work on Agere FW643-E (same vendor and device ID, but rev 07),\nintroduce a device revision field into firewire-ohci\u0027s quirks list so\nthat different quirks can be defined for older and newer revisions.\n\nReported-by: Jonathan Isom \u003cjeisom@gmail.com\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\nCc: \u003cstable@kernel.org\u003e # 2.6.36.y\n" }, { "commit": "af0cdf4947818becfe209610b209315578645ab4", "tree": "6294d93f9eccfb857769929d1269aedd32681fc9", "parents": [ "cf7d7e5a1980d1116ee152d25dac382b112b9c17" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Dec 07 19:16:02 2010 +0100" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Dec 12 15:47:02 2010 +0100" }, "message": "firewire: ohci: fix regression with VIA VT6315, disable MSI\n\n\"VIA Technologies, Inc. VT6315 Series Firewire Controller [1106:3403]\"\ndoes not generate any interrupts if Message Signaled Interrupts were\nenabled. This is a regression since kernel 2.6.36 in which MSI support\nwas added to firewire-ohci. Hence blacklist MSI on all VIA controllers.\n\nReported-by: Robin Cook \u003crcook@wyrms.net\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\nCc: \u003cstable@kernel.org\u003e # 2.6.36.y\n" }, { "commit": "75f5d2c9bd36047364d1fb35c0720ab37df3be02", "tree": "20f7bfba3397366e0f4526dcdbdb20f23559d8a5", "parents": [ "0f639a3c5ca63dd76ee07de9b02ebf0178ce9a17", "b2268830f5cf29d94b3e4a2af0b795a8f28776fe" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Nov 28 12:24:20 2010 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Nov 28 12:24:20 2010 -0800" }, "message": "Merge branch \u0027fwnet\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6\n\n* \u0027fwnet\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:\n firewire: net: throttle TX queue before running out of tlabels\n firewire: net: replace lists by counters\n firewire: net: fix memory leaks\n firewire: net: count stats.tx_packets and stats.tx_bytes\n" }, { "commit": "b2268830f5cf29d94b3e4a2af0b795a8f28776fe", "tree": "a52483756446f1e09dcd80b5243d899612cf5caa", "parents": [ "48553011cea504796e513350740781ac6745f556" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Nov 14 14:35:40 2010 +0100" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Nov 17 00:08:49 2010 +0100" }, "message": "firewire: net: throttle TX queue before running out of tlabels\n\nThis prevents firewire-net from submitting write requests in fast\nsuccession until failure due to all 64 transaction labels were used up\nfor unfinished split transactions. The netif_stop/wake_queue API is\nused for this purpose.\n\nWithout this stop/wake mechanism, datagrams were simply lost whenever\nthe tlabel pool was exhausted. Plus, tlabel exhaustion by firewire-net\nalso prevented other unrelated outbound transactions to be initiated.\n\nThe chosen queue depth was checked by me to hit the maximum possible\nthroughput with an OS X peer whose receive DMA is good enough to never\nreject requests due to busy inbound request FIFO. Current Linux peers\nshow a mixed picture of -5%...+15% change in bandwidth; their current\nbottleneck are RCODE_BUSY situations (fewer or more, depending on TX\nqueue depth) due to too small AR buffer in firewire-ohci.\n\nMaxim Levitsky tested this change with similar watermarks with a Linux\npeer and some pending firewire-ohci improvements that address the\nRCODE_BUSY problem and confirmed that these TX queue limits are good.\n\nNote: This removes some netif_wake_queue from reception code paths.\nThey were apparently copy\u0026paste artefacts from a nonsensical\nnetif_wake_queue use in the older eth1394 driver. This belongs only\ninto the transmit path.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\nTested-by: Maxim Levitsky \u003cmaximlevitsky@gmail.com\u003e\n" }, { "commit": "48553011cea504796e513350740781ac6745f556", "tree": "f90a9e23ecd00c9c476e61e6e2c2fdd6cc552008", "parents": [ "7ee11fa8d0a84b05cefe12b0bebc05ab0ea89cd6" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Nov 06 23:18:23 2010 +0100" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Nov 17 00:08:48 2010 +0100" }, "message": "firewire: net: replace lists by counters\n\nThe current transmit code does not at all make use of\n - fwnet_device.packet_list\nand only very limited use of\n - fwnet_device.broadcasted_list,\n - fwnet_device.queued_packets.\nTheir current function is to track whether the TX soft-IRQ finished\ndealing with an skb when the AT-req tasklet takes over, and to discard\npending tx datagrams (if there are any) when the local node is removed.\n\nThe latter does actually contain a race condition bug with TX soft-IRQ\nand AT-req tasklet.\n\nInstead of these lists and the corresponding link in fwnet_packet_task,\n - a flag in fwnet_packet_task to track whether fwnet_tx is done,\n - a counter of queued datagrams in fwnet_device\ndo the job as well.\n\nThe above mentioned theoretic race condition is resolved by letting\nfwnet_remove sleep until all datagrams were flushed. It may sleep\nalmost arbitrarily long since fwnet_remove is executed in the context of\na multithreaded (concurrency managed) workqueue.\n\nThe type of max_payload is changed to u16 here to avoid waste in struct\nfwnet_packet_task. This value cannot exceed 4096 per IEEE 1394:2008\ntable 16-18 (or 32678 per specification of packet headers, if there is\never going to be something else than beta mode).\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "7ee11fa8d0a84b05cefe12b0bebc05ab0ea89cd6", "tree": "ed0f67e8b3687aef93acd50db1f635d43c8073f5", "parents": [ "902bca00dc6e3b3ff5fbb1e32e5dbb45d5f30579" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Nov 06 16:57:28 2010 +0100" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Nov 17 00:08:48 2010 +0100" }, "message": "firewire: net: fix memory leaks\n\na) fwnet_transmit_packet_done used to poison ptask-\u003ept_link by list_del.\nIf fwnet_send_packet checked later whether it was responsible to clean\nup (in the border case that the TX soft IRQ was outpaced by the AT-req\ntasklet on another CPU), it missed this because ptask-\u003ept_link was no\nlonger shown as empty.\n\nb) If fwnet_write_complete got an rcode other than RCODE_COMPLETE, we\nmissed to free the skb and ptask entirely.\n\nAlso, count stats.tx_dropped and stats.tx_errors when rcode !\u003d 0.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "902bca00dc6e3b3ff5fbb1e32e5dbb45d5f30579", "tree": "4800783df10844fc3f41071fe4446cea957f63e4", "parents": [ "e53beacd23d9cb47590da6a7a7f6d417b941a994" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Nov 06 12:36:13 2010 +0100" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Nov 17 00:08:48 2010 +0100" }, "message": "firewire: net: count stats.tx_packets and stats.tx_bytes\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "f281233d3eba15fb225d21ae2e228fd4553d824a", "tree": "51134454ba8acb558735f90be5540f7d756483e3", "parents": [ "bdbd01ac444bffb3c9aefed3059d12554059b320" ], "author": { "name": "Jeff Garzik", "email": "jeff@garzik.org", "time": "Tue Nov 16 02:10:29 2010 -0500" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 16 13:33:23 2010 -0800" }, "message": "SCSI host lock push-down\n\nMove the mid-layer\u0027s -\u003equeuecommand() invocation from being locked\nwith the host lock to being unlocked to facilitate speeding up the\ncritical path for drivers who don\u0027t need this lock taken anyway.\n\nThe patch below presents a simple SCSI host lock push-down as an\nequivalent transformation. No locking or other behavior should change\nwith this patch. All existing bugs and locking orders are preserved.\n\nAdditionally, add one parameter to queuecommand,\n\tstruct Scsi_Host *\nand remove one parameter from queuecommand,\n\tvoid (*done)(struct scsi_cmnd *)\n\nScsi_Host* is a convenient pointer that most host drivers need anyway,\nand \u0027done\u0027 is redundant to struct scsi_cmnd-\u003escsi_done.\n\nMinimal code disturbance was attempted with this change. Most drivers\nneeded only two one-line modifications for their host lock push-down.\n\nSigned-off-by: Jeff Garzik \u003cjgarzik@redhat.com\u003e\nAcked-by: James Bottomley \u003cJames.Bottomley@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "f69fa76482e654f7d94e4aa40ea0ebf04363396a", "tree": "3cb1976afa6daf39a9e45551671ecfb5bb580e08", "parents": [ "2e5c36722d4c9c86281f25a1e963a6078c7fce6a", "693fa7792e9db9f32da9436e633976fbacd04b55" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Nov 05 14:17:22 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Nov 05 14:17:22 2010 -0700" }, "message": "Merge branch \u0027fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6\n\n* \u0027fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:\n firewire: ohci: fix race when reading count in AR descriptor\n firewire: ohci: avoid reallocation of AR buffers\n firewire: ohci: fix race in AR split packet handling\n firewire: ohci: fix buffer overflow in AR split packet handling\n" }, { "commit": "693fa7792e9db9f32da9436e633976fbacd04b55", "tree": "227078047db9c8f3497133769b1b9b1f7b681fa1", "parents": [ "837596a61ba8f9bb53bb7aa27d17328ff9b2bcd5" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Mon Oct 25 11:43:05 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Oct 30 23:37:20 2010 +0200" }, "message": "firewire: ohci: fix race when reading count in AR descriptor\n\nIf the controller is storing a split packet and therefore changing\nd-\u003eres_count to zero between the two reads by the driver, we end up with\nan end pointer that is not at a packet boundary, and therefore overflow\nthe buffer when handling the split packet.\n\nTo fix this, read the field once, atomically. The compiler usually\nmerges the two reads anyway, but for correctness, we have to enforce it.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nTested-by: Maxim Levitsky \u003cmaximlevitsky@gmail.com\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "837596a61ba8f9bb53bb7aa27d17328ff9b2bcd5", "tree": "a9b2ba0d4382a1c0ad75182662570f61b5e03e57", "parents": [ "a1f805e5e73a8fe166b71c6592d3837df0cd5e2e" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Mon Oct 25 11:42:42 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Oct 30 23:37:20 2010 +0200" }, "message": "firewire: ohci: avoid reallocation of AR buffers\n\nFreeing an AR buffer page just to allocate a new page immediately\nafterwards is not only a pointless effort but also dangerous because\nthe allocation can fail, which would result in an oops later.\n\nSplit ar_context_add_page() into two functions so that we can reuse\nthe old page directly.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nTested-by: Maxim Levitsky \u003cmaximlevitsky@gmail.com\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "a1f805e5e73a8fe166b71c6592d3837df0cd5e2e", "tree": "f8bc4a2d8d0bcaf4a1b7d4c547d7a2525459f852", "parents": [ "85f7ffd5d2b320f73912b15fe8cef34bae297daf" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Mon Oct 25 11:42:20 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Oct 30 23:37:19 2010 +0200" }, "message": "firewire: ohci: fix race in AR split packet handling\n\nWhen handling an AR buffer that has been completely filled, we assumed\nthat its descriptor will not be read by the controller and can be\noverwritten. However, when the last received packet happens to end at\nthe end of the buffer, the controller might not yet have moved on to the\nnext buffer and might read the branch address later. If we overwrite\nand free the page before that, the DMA context will either go dead\nbecause of an invalid Z value, or go off into some random memory.\n\nTo fix this, ensure that the descriptor does not get overwritten by\nusing only the actual buffer instead of the entire page for reassembling\nthe split packet. Furthermore, to avoid freeing the page too early,\nmove on to the next buffer only when some data in it guarantees that the\ncontroller has moved on.\n\nThis should eliminate the remaining firewire-net problems.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nCc: 2.6.22-2.6.36 \u003cstable@kernel.org\u003e\nTested-by: Maxim Levitsky \u003cmaximlevitsky@gmail.com\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "85f7ffd5d2b320f73912b15fe8cef34bae297daf", "tree": "d9151979bb3834bd1254239a42247a668f059133", "parents": [ "f6f94e2ab1b33f0082ac22d71f66385a60d8157f" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Mon Oct 25 11:41:53 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Oct 30 23:37:19 2010 +0200" }, "message": "firewire: ohci: fix buffer overflow in AR split packet handling\n\nWhen the controller had to split a received asynchronous packet into two\nbuffers, the driver tries to reassemble it by copying both parts into\nthe first page. However, if size + rest \u003e PAGE_SIZE, i.e., if the yet\nunhandled packets before the split packet, the split packet itself, and\nany received packets after the split packet are together larger than one\npage, then the memory after the first page would get overwritten.\n\nTo fix this, do not try to copy the data of all unhandled packets at\nonce, but copy the possibly needed data every time when handling\na packet.\n\nThis gets rid of most of the infamous crashes and data corruptions when\nusing firewire-net.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nCc: 2.6.22-2.6.36 \u003cstable@kernel.org\u003e\nTested-by: Maxim Levitsky \u003cmaximlevitsky@gmail.com\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e (cast PAGE_SIZE to size_t)\n" }, { "commit": "b7d41a9fbb364c67d91c3588e117eba547e2d4bf", "tree": "abf95ce77eaa8b7aed30fded4d3592b4af152a99", "parents": [ "da47c19e5c746829042933c8f945a71e2b62d6fc", "66fa12c571d35e3cd62574c65f1785a460105397" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Oct 25 08:05:29 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Oct 25 08:05:29 2010 -0700" }, "message": "Merge branch \u0027ieee1394-removal\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6\n\n* \u0027ieee1394-removal\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:\n ieee1394: remove the old IEEE 1394 driver stack\n ieee1394: move init_ohci1394_dma to drivers/firewire/\n\nFix trivial change/delete conflict: drivers/ieee1394/eth1394.c is\ngetting removed, but was modified by the networking merge.\n" }, { "commit": "5f05647dd81c11a6a165ccc8f0c1370b16f3bcb0", "tree": "7851ef1c93aa1aba7ef327ca4b75fd35e6d10f29", "parents": [ "02f36038c568111ad4fc433f6fa760ff5e38fab4", "ec37a48d1d16c30b655ac5280209edf52a6775d4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 23 11:47:02 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Oct 23 11:47:02 2010 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6: (1699 commits)\n bnx2/bnx2x: Unsupported Ethtool operations should return -EINVAL.\n vlan: Calling vlan_hwaccel_do_receive() is always valid.\n tproxy: use the interface primary IP address as a default value for --on-ip\n tproxy: added IPv6 support to the socket match\n cxgb3: function namespace cleanup\n tproxy: added IPv6 support to the TPROXY target\n tproxy: added IPv6 socket lookup function to nf_tproxy_core\n be2net: Changes to use only priority codes allowed by f/w\n tproxy: allow non-local binds of IPv6 sockets if IP_TRANSPARENT is enabled\n tproxy: added tproxy sockopt interface in the IPV6 layer\n tproxy: added udp6_lib_lookup function\n tproxy: added const specifiers to udp lookup functions\n tproxy: split off ipv6 defragmentation to a separate module\n l2tp: small cleanup\n nf_nat: restrict ICMP translation for embedded header\n can: mcp251x: fix generation of error frames\n can: mcp251x: fix endless loop in interrupt handler if CANINTF_MERRF is set\n can-raw: add msg_flags to distinguish local traffic\n 9p: client code cleanup\n rds: make local functions/variables static\n ...\n\nFix up conflicts in net/core/dev.c, drivers/net/pcmcia/smc91c92_cs.c and\ndrivers/net/wireless/ath/ath9k/debug.c as per David\n" }, { "commit": "092e0e7e520a1fca03e13c9f2d157432a8657ff2", "tree": "451897252c4c08c4b5a8ef535da156f1e817e80b", "parents": [ "79f14b7c56d3b3ba58f8b43d1f70b9b71477a800", "776c163b1b93c8dfa5edba885bc2bfbc2d228a5f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 22 10:52:56 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Oct 22 10:52:56 2010 -0700" }, "message": "Merge branch \u0027llseek\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/bkl\n\n* \u0027llseek\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/bkl:\n vfs: make no_llseek the default\n vfs: don\u0027t use BKL in default_llseek\n llseek: automatically add .llseek fop\n libfs: use generic_file_llseek for simple_attr\n mac80211: disallow seeks in minstrel debug code\n lirc: make chardev nonseekable\n viotape: use noop_llseek\n raw: use explicit llseek file operations\n ibmasmfs: use generic_file_llseek\n spufs: use llseek in all file operations\n arm/omap: use generic_file_llseek in iommu_debug\n lkdtm: use generic_file_llseek in debugfs\n net/wireless: use generic_file_llseek in debugfs\n drm: use noop_llseek\n" }, { "commit": "2198a10b501fd4443430cb17e065a9e859cc58c9", "tree": "87f3781d293da0f8f8f61615905eb7bf62b7c128", "parents": [ "9941fb62762253774cc6177d0b9172ece5133fe1", "db5a753bf198ef7a50e17d2ff358adf37efe8648" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Oct 21 08:43:05 2010 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Oct 21 08:43:05 2010 -0700" }, "message": "Merge branch \u0027master\u0027 of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6\n\nConflicts:\n\tnet/core/dev.c\n" }, { "commit": "aa0170fff3c26bf2b42159af2dd9cf86444c292a", "tree": "0a64674f2fa51883c66e96c743dd4c4e86d98a49", "parents": [ "cd07202cc8262e1669edff0d97715f3dd9260917" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Oct 17 14:09:12 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Oct 17 14:09:12 2010 +0200" }, "message": "firewire: ohci: fix TI TSB82AA2 regression since 2.6.35\n\nRevert commit 54672386ccf36ffa21d1de8e75624af83f9b0eeb\n\"firewire: ohci: fix up configuration of TI chips\".\nIt caused massive slow-down and data corruption with a TSB82AA2 based\nStarTech EC1394B2 ExpressCard and FireWire 800 harddisks.\n\nhttps://bugs.launchpad.net/ubuntu/+source/linux/+bug/657081\nhttp://thread.gmane.org/gmane.linux.kernel.firewire.user/4013\n\nThe fact that some card EEPROMs do not program these enhancements may be\nrelated to TSB81BA3 phy chip errata, if not to bugs of TSB82AA2 itself.\nWe could re-add these configuration steps, but only conditional on a\nwhitelist of cards on which these enhancements bring a proven positive\neffect.\n\nReported-and-tested-by: Eric Shattow \u003clucent@gmail.com\u003e\nCc: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nCc: \u003cstable@kernel.org\u003e 2.6.35\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "6038f373a3dc1f1c26496e60b6c40b164716f07e", "tree": "a0d3bbd026eea41b9fc36b8c722cbaf56cd9f825", "parents": [ "1ec5584e3edf9c4bf2c88c846534d19cf986ba11" ], "author": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Sun Aug 15 18:52:59 2010 +0200" }, "committer": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Fri Oct 15 15:53:27 2010 +0200" }, "message": "llseek: automatically add .llseek fop\n\nAll file_operations should get a .llseek operation so we can make\nnonseekable_open the default for future file operations without a\n.llseek pointer.\n\nThe three cases that we can automatically detect are no_llseek, seq_lseek\nand default_llseek. For cases where we can we can automatically prove that\nthe file offset is always ignored, we use noop_llseek, which maintains\nthe current behavior of not returning an error from a seek.\n\nNew drivers should normally not use noop_llseek but instead use no_llseek\nand call nonseekable_open at open time. Existing drivers can be converted\nto do the same when the maintainer knows for certain that no user code\nrelies on calling seek on the device file.\n\nThe generated code is often incorrectly indented and right now contains\ncomments that clarify for each added line why a specific variant was\nchosen. In the version that gets submitted upstream, the comments will\nbe gone and I will manually fix the indentation, because there does not\nseem to be a way to do that using coccinelle.\n\nSome amount of new code is currently sitting in linux-next that should get\nthe same modifications, which I will do at the end of the merge window.\n\nMany thanks to Julia Lawall for helping me learn to write a semantic\npatch that does all this.\n\n\u003d\u003d\u003d\u003d\u003d begin semantic patch \u003d\u003d\u003d\u003d\u003d\n// This adds an llseek\u003d method to all file operations,\n// as a preparation for making no_llseek the default.\n//\n// The rules are\n// - use no_llseek explicitly if we do nonseekable_open\n// - use seq_lseek for sequential files\n// - use default_llseek if we know we access f_pos\n// - use noop_llseek if we know we don\u0027t access f_pos,\n// but we still want to allow users to call lseek\n//\n@ open1 exists @\nidentifier nested_open;\n@@\nnested_open(...)\n{\n\u003c+...\nnonseekable_open(...)\n...+\u003e\n}\n\n@ open exists@\nidentifier open_f;\nidentifier i, f;\nidentifier open1.nested_open;\n@@\nint open_f(struct inode *i, struct file *f)\n{\n\u003c+...\n(\nnonseekable_open(...)\n|\nnested_open(...)\n)\n...+\u003e\n}\n\n@ read disable optional_qualifier exists @\nidentifier read_f;\nidentifier f, p, s, off;\ntype ssize_t, size_t, loff_t;\nexpression E;\nidentifier func;\n@@\nssize_t read_f(struct file *f, char *p, size_t s, loff_t *off)\n{\n\u003c+...\n(\n *off \u003d E\n|\n *off +\u003d E\n|\n func(..., off, ...)\n|\n E \u003d *off\n)\n...+\u003e\n}\n\n@ read_no_fpos disable optional_qualifier exists @\nidentifier read_f;\nidentifier f, p, s, off;\ntype ssize_t, size_t, loff_t;\n@@\nssize_t read_f(struct file *f, char *p, size_t s, loff_t *off)\n{\n... when !\u003d off\n}\n\n@ write @\nidentifier write_f;\nidentifier f, p, s, off;\ntype ssize_t, size_t, loff_t;\nexpression E;\nidentifier func;\n@@\nssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off)\n{\n\u003c+...\n(\n *off \u003d E\n|\n *off +\u003d E\n|\n func(..., off, ...)\n|\n E \u003d *off\n)\n...+\u003e\n}\n\n@ write_no_fpos @\nidentifier write_f;\nidentifier f, p, s, off;\ntype ssize_t, size_t, loff_t;\n@@\nssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off)\n{\n... when !\u003d off\n}\n\n@ fops0 @\nidentifier fops;\n@@\nstruct file_operations fops \u003d {\n ...\n};\n\n@ has_llseek depends on fops0 @\nidentifier fops0.fops;\nidentifier llseek_f;\n@@\nstruct file_operations fops \u003d {\n...\n .llseek \u003d llseek_f,\n...\n};\n\n@ has_read depends on fops0 @\nidentifier fops0.fops;\nidentifier read_f;\n@@\nstruct file_operations fops \u003d {\n...\n .read \u003d read_f,\n...\n};\n\n@ has_write depends on fops0 @\nidentifier fops0.fops;\nidentifier write_f;\n@@\nstruct file_operations fops \u003d {\n...\n .write \u003d write_f,\n...\n};\n\n@ has_open depends on fops0 @\nidentifier fops0.fops;\nidentifier open_f;\n@@\nstruct file_operations fops \u003d {\n...\n .open \u003d open_f,\n...\n};\n\n// use no_llseek if we call nonseekable_open\n////////////////////////////////////////////\n@ nonseekable1 depends on !has_llseek \u0026\u0026 has_open @\nidentifier fops0.fops;\nidentifier nso ~\u003d \"nonseekable_open\";\n@@\nstruct file_operations fops \u003d {\n... .open \u003d nso, ...\n+.llseek \u003d no_llseek, /* nonseekable */\n};\n\n@ nonseekable2 depends on !has_llseek @\nidentifier fops0.fops;\nidentifier open.open_f;\n@@\nstruct file_operations fops \u003d {\n... .open \u003d open_f, ...\n+.llseek \u003d no_llseek, /* open uses nonseekable */\n};\n\n// use seq_lseek for sequential files\n/////////////////////////////////////\n@ seq depends on !has_llseek @\nidentifier fops0.fops;\nidentifier sr ~\u003d \"seq_read\";\n@@\nstruct file_operations fops \u003d {\n... .read \u003d sr, ...\n+.llseek \u003d seq_lseek, /* we have seq_read */\n};\n\n// use default_llseek if there is a readdir\n///////////////////////////////////////////\n@ fops1 depends on !has_llseek \u0026\u0026 !nonseekable1 \u0026\u0026 !nonseekable2 \u0026\u0026 !seq @\nidentifier fops0.fops;\nidentifier readdir_e;\n@@\n// any other fop is used that changes pos\nstruct file_operations fops \u003d {\n... .readdir \u003d readdir_e, ...\n+.llseek \u003d default_llseek, /* readdir is present */\n};\n\n// use default_llseek if at least one of read/write touches f_pos\n/////////////////////////////////////////////////////////////////\n@ fops2 depends on !fops1 \u0026\u0026 !has_llseek \u0026\u0026 !nonseekable1 \u0026\u0026 !nonseekable2 \u0026\u0026 !seq @\nidentifier fops0.fops;\nidentifier read.read_f;\n@@\n// read fops use offset\nstruct file_operations fops \u003d {\n... .read \u003d read_f, ...\n+.llseek \u003d default_llseek, /* read accesses f_pos */\n};\n\n@ fops3 depends on !fops1 \u0026\u0026 !fops2 \u0026\u0026 !has_llseek \u0026\u0026 !nonseekable1 \u0026\u0026 !nonseekable2 \u0026\u0026 !seq @\nidentifier fops0.fops;\nidentifier write.write_f;\n@@\n// write fops use offset\nstruct file_operations fops \u003d {\n... .write \u003d write_f, ...\n+\t.llseek \u003d default_llseek, /* write accesses f_pos */\n};\n\n// Use noop_llseek if neither read nor write accesses f_pos\n///////////////////////////////////////////////////////////\n\n@ fops4 depends on !fops1 \u0026\u0026 !fops2 \u0026\u0026 !fops3 \u0026\u0026 !has_llseek \u0026\u0026 !nonseekable1 \u0026\u0026 !nonseekable2 \u0026\u0026 !seq @\nidentifier fops0.fops;\nidentifier read_no_fpos.read_f;\nidentifier write_no_fpos.write_f;\n@@\n// write fops use offset\nstruct file_operations fops \u003d {\n...\n .write \u003d write_f,\n .read \u003d read_f,\n...\n+.llseek \u003d noop_llseek, /* read and write both use no f_pos */\n};\n\n@ depends on has_write \u0026\u0026 !has_read \u0026\u0026 !fops1 \u0026\u0026 !fops2 \u0026\u0026 !has_llseek \u0026\u0026 !nonseekable1 \u0026\u0026 !nonseekable2 \u0026\u0026 !seq @\nidentifier fops0.fops;\nidentifier write_no_fpos.write_f;\n@@\nstruct file_operations fops \u003d {\n... .write \u003d write_f, ...\n+.llseek \u003d noop_llseek, /* write uses no f_pos */\n};\n\n@ depends on has_read \u0026\u0026 !has_write \u0026\u0026 !fops1 \u0026\u0026 !fops2 \u0026\u0026 !has_llseek \u0026\u0026 !nonseekable1 \u0026\u0026 !nonseekable2 \u0026\u0026 !seq @\nidentifier fops0.fops;\nidentifier read_no_fpos.read_f;\n@@\nstruct file_operations fops \u003d {\n... .read \u003d read_f, ...\n+.llseek \u003d noop_llseek, /* read uses no f_pos */\n};\n\n@ depends on !has_read \u0026\u0026 !has_write \u0026\u0026 !fops1 \u0026\u0026 !fops2 \u0026\u0026 !has_llseek \u0026\u0026 !nonseekable1 \u0026\u0026 !nonseekable2 \u0026\u0026 !seq @\nidentifier fops0.fops;\n@@\nstruct file_operations fops \u003d {\n...\n+.llseek \u003d noop_llseek, /* no read or write fn */\n};\n\u003d\u003d\u003d\u003d\u003d End semantic patch \u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nCc: Julia Lawall \u003cjulia@diku.dk\u003e\nCc: Christoph Hellwig \u003chch@infradead.org\u003e\n" }, { "commit": "66fa12c571d35e3cd62574c65f1785a460105397", "tree": "b4f8de3d5ca827d2b134ed628628a7bff46967ca", "parents": [ "1ef5b816c0eaf84f91106cfc0893069c49e86113" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Oct 10 00:12:20 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Oct 11 14:48:03 2010 +0200" }, "message": "ieee1394: remove the old IEEE 1394 driver stack\n\nThe drivers\n - ohci1394 (controller driver)\n - ieee1394 (core)\n - dv1394, raw1394, video1394 (userspace ABI)\n - eth1394, sbp2 (protocol drivers)\nare replaced by\n - firewire-ohci (controller driver)\n - firewire-core (core and userspace ABI)\n - firewire-net, firewire-sbp2 (protocol drivers)\nwhich are more featureful, better performing, and more secure than the older\ndrivers; all with a smaller and more modern code base.\n\nThe driver firedtv in drivers/media/dvb/firewire/ contains backends to both\nieee1394 and firewire-core. Its ieee1394 backend code can be removed in an\nindependent commit; firedtv as-is builds and works fine without ieee1394.\n\nThe driver pcilynx (an incomplete controller driver) is deleted without\nreplacement since PCILynx cards are extremely rare. Owners of these cards\nuse them with the stand-alone bus sniffer driver nosy instead.\n\nThe drivers nosy and init_ohci1394_dma which do not interact with either of\nthe two IEEE 1394 stacks are not affected by the ieee1394 subsystem removal.\n\nThere are still some issues with the newer firewire subsystem compared to\nthe older one:\n - The rare and quirky controllers ALi M52xx, Apple UniNorth v1, NVIDIA\n NForce2 are even less well supported by firewire-ohci than by ohci1394.\n I am looking into the M52xx issue.\n - The experimental firewire-net is reportedly less stable than its\n experimental cousin eth1394.\n - Audio playback of a certain group of audio devices (ones based on DICE\n chipset with EAP; supported by prerelease FFADO code) does not work yet.\n This issue is still under investigation.\n - There were some ieee1394 based out-of-the-mainline drivers. Of them,\n only lisight, an audio driver for iSight webcams, seems still useful.\n Work is underway to reimplement it on top of firewire-core.\n\nAll these remainig issues are minor; they should not stand in the way of\noverall better user experience of IEEE 1394 on Linux, together with a\nreduction in support efforts and maintenance burden. The coexistence of two\nIEEE 1394 kernel driver stacks in the mainline since 2.6.22 shall end now,\nas announced earlier this year.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "1ef5b816c0eaf84f91106cfc0893069c49e86113", "tree": "5db2c91a18ce2aa40177c2c968b0ed5f15de251b", "parents": [ "cb655d0f3d57c23db51b981648e452988c0223f9" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Oct 10 00:54:02 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Oct 11 14:48:03 2010 +0200" }, "message": "ieee1394: move init_ohci1394_dma to drivers/firewire/\n\nbecause drivers/ieee1394/ will be deleted.\n\nAdditional changes:\n - add some #include directives\n - adjust to use firewire/ohci.h instead of ieee1394/ohci1394.h,\n replace struct ti_ohci by a minimal struct ohci,\n replace quadlet_t from ieee1394_types.h by u32\n - two or three trivial stylistic changes\n - __iomem annotation\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "69259abb64d4da77273bf59accfc9fa79e7165f4", "tree": "bd043ab03a788b749c8d5ae4049d8defae9abf34", "parents": [ "dd53df265b1ee7a1fbbc76bb62c3bec2383bbd44", "12e94471b2be5ef9b55b10004a3a2cd819490036" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Oct 06 19:39:31 2010 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Oct 06 19:39:31 2010 -0700" }, "message": "Merge branch \u0027master\u0027 of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6\n\nConflicts:\n\tdrivers/net/pcmcia/pcnet_cs.c\n\tnet/caif/caif_socket.c\n" }, { "commit": "e548833df83c3554229eff0672900bfe958b45fd", "tree": "85efc4a76dc356593d6d394776aeb845dc580fb6", "parents": [ "cbd9da7be869f676afc204e1a664163778c770bd", "053d8f6622701f849fda2ca2c9ae596c13599ba9" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 09 22:27:33 2010 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 09 22:27:33 2010 -0700" }, "message": "Merge branch \u0027master\u0027 of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6\n\nConflicts:\n\tnet/mac80211/main.c\n" }, { "commit": "970f4be85ae6ecf97b711a3a2a1d5cecd3ea0534", "tree": "4088ff5bf8d3d020ddfe53fdb6d4252ff3457914", "parents": [ "a4dc090b6cb445257d2a8e44f85395ced6d1ed3e" ], "author": { "name": "Heikki Lindholm", "email": "holin@iki.fi", "time": "Mon Sep 06 22:30:45 2010 +0300" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Sep 08 21:25:55 2010 +0200" }, "message": "firewire: ohci: activate cycle timer register quirk on Ricoh chips\n\nThe Ricoh FireWire controllers appear to have the non-atomic cycle\ntimer register access bug, so, activate the driver workaround by\ndefault.\n\nThe behaviour was observed on:\nRicoh Co Ltd R5C552 IEEE 1394 Controller [1180:0552] and\nRicoh Co Ltd R5C832 IEEE 1394 Controller [1180:0832] (rev 04).\n\nSigned-off-by: Heikki Lindholm \u003cholin@iki.fi\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "a4dc090b6cb445257d2a8e44f85395ced6d1ed3e", "tree": "1cba172819b1395237adb69e34dfa469a63dbb68", "parents": [ "2222bcb76790f4f61f39ec1514946a7593b07e02" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Aug 28 14:21:26 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Aug 29 09:17:31 2010 +0200" }, "message": "firewire: ohci: work around VIA and NEC PHY packet reception bug\n\nVIA VT6306, VIA VT6308, and NEC OrangeLink controllers do not write\npacket event codes for received PHY packets (or perhaps write\nevt_no_status, hard to tell). Work around it by overwriting the\npacket\u0027s ACK by ack_complete, so that upper layers that listen to PHY\npacket reception get to see these packets.\n\n(Also tested: TI TSB82AA2, TI TSB43AB22/A, TI XIO2213A, Agere FW643,\nJMicron JMB381 --- these do not exhibit this bug.)\n\nClemens proposed a quirks flag for that, IOW whitelist known misbehaving\ncontrollers for this workaround. Though to me it seems harmless enough\nto enable for all controllers.\n\nThe log_ar_at_event() debug log will continue to show the original\nstatus from the DMA unit.\n\nReported-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e (VT6308)\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "2222bcb76790f4f61f39ec1514946a7593b07e02", "tree": "f479ec53c083071d70935b51eb7adab844233779", "parents": [ "1bf145fed572583d4cb7c1784689a0b42c997ba6" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Wed Aug 18 15:05:02 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Aug 19 20:28:25 2010 +0200" }, "message": "firewire: core: do not use del_timer_sync() in interrupt context\n\nBecause we might be in interrupt context, replace del_timer_sync() with\ndel_timer(). If the timer is already running, we know that it will\nclean up the transaction, so we do not need to do any further processing\nin the normal transaction handler.\n\nMany thanks to Yong Zhang for diagnosing this.\n\nReported-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "1bf145fed572583d4cb7c1784689a0b42c997ba6", "tree": "3ef524b2af2ea77f2ea99478b8a4058a8cf07d13", "parents": [ "a481e97d3cdc40b9d58271675bd4f0abb79d4872" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Aug 16 23:45:54 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Aug 19 20:28:25 2010 +0200" }, "message": "firewire: net: fix unicast reception RCODE in failure paths\n\nThe incoming request hander fwnet_receive_packet() expects subsequent\ndatagram handling code to return non-zero on errors. However, almost\nnone of the failure paths did so. Fix them all.\n\n(This error reporting is used to send and RCODE_CONFLICT_ERROR to the\nsender node in such failure cases. Two modes of failure exist: Out of\nmemory, or firewire-net is unaware of any peer node to which a fragment\nor an ARP packet belongs. However, it is unclear whether a sender can\nactually make use of such information. A Linux peer apparently can\u0027t.\nMaybe it should all be simplified to void functions.)\n\nReported-by: Julia Lawall \u003cjulia@diku.dk\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "a481e97d3cdc40b9d58271675bd4f0abb79d4872", "tree": "c860d626fbee5f1fb2bc9511fa7fb3c586801c56", "parents": [ "6c74340bce253ea95c9ee801b3c411a333937edf" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Aug 16 22:13:34 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Aug 19 20:28:25 2010 +0200" }, "message": "firewire: sbp2: fix stall with \"Unsolicited response\"\n\nFix I/O stalls with some 4-bay RAID enclosures which are based on\nOXUF936QSE:\n - Onnto dataTale RSM4QO, old firmware (not anymore with current\n firmware),\n - inXtron Hydra Super-S LCM, old as well as current firmware\nwhen used in RAID-5 mode, perhaps also in other RAID modes.\n\nThe stalls happen during heavy or moderate disk traffic in periods that\nare a multiple of 5 minutes, roughly twice per hour. They are caused\nby the target responding too late to an ORB_Pointer register write:\nThe target responds after Split_Timeout, hence firewire-core cancels\nthe transaction, and firewire-sbp2 fails the SCSI request. The SCSI\ncore retries the request, that fails again (and again), hence SCSI core\ncalls firewire-sbp2\u0027s abort handler (and even the Management_Agent\nregister write in the abort handler has the transaction timeout\nproblem).\n\nDuring all that, the process which issued the I/O is stalled in I/O\nwait state.\n\nMeanwhile, the target actually acts on the first failed SCSI request:\nIt responds to the ORB_Pointer write later (seen in the kernel log as\n\"firewire_core: Unsolicited response\") and also finishes the SCSI\nrequest with proper status (seen in the kernel log as \"firewire_sbp2:\nstatus write for unknown orb\").\n\nSo let\u0027s just ignore RCODE_CANCELLED in the transaction callback and\nwait for the target to complete the ORB nevertheless. This requires\na small modification is sbp2_cancel_orbs(); it now needs to call\norb-\u003ecallback() regardless whether fw_cancel_transaction() found the\ntransaction unfinished or finished.\n\nA different solution is to increase Split_Timeout on the local node.\n(Tested: 2000ms timeout; maybe 1000ms or something like that works too.\n200ms is insufficient. Standard is 100ms.) However, I rather not do\nthis because any software on any node could change the Split_Timeout to\nsomething unsuitable. Or such a large Split_Timeout may be undesirable\nfor other purposes.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "6c74340bce253ea95c9ee801b3c411a333937edf", "tree": "4be487165cfc295cba5613094430248f6fb968e9", "parents": [ "840fe6359c1db978d01fceb8a023f5f6efdf7f1c" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Aug 16 21:58:03 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Aug 19 20:28:25 2010 +0200" }, "message": "firewire: sbp2: fix memory leak in sbp2_cancel_orbs or at send error\n\nWhen an ORB was canceled (Command ORB i.e. SCSI request timed out, or\nManagement ORB timed out), or there was a send error in the initial\ntransaction, we missed to drop one of the ORB\u0027s references and thus\nleaked memory.\n\nBackground:\nIn total, we hold 3 references to each Operation Request Block:\n - 1 during sbp2_scsi_queuecommand() or sbp2_send_management_orb()\n respectively,\n - 1 for the duration of the write transaction to the ORB_Pointer or\n Management_Agent register of the target,\n - 1 for as long as the ORB stays within the lu-\u003eorb_list, until\n the ORB is unlinked from the list and the orb-\u003ecallback was\n executed.\n\nThe latter one of these 3 references is finished\n - normally by sbp2_status_write() when the target wrote status\n for a pending ORB,\n - or by sbp2_cancel_orbs() in case of an ORB time-out,\n - or by complete_transaction() in case of a send error.\nOf them, the latter two lacked the kref_put.\n\nAdd the missing kref_put()s. Add comments to the gets and puts of\nreferences for transaction callbacks and ORB callbacks so that it is\neasier to see what is supposed to happen.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "01414802054c382072b6cb9a1bdc6e243c74b2d5", "tree": "f09b05eb9fff16a69d83ec28653992d1bae8d74c", "parents": [ "bb8a10bbd10a45db0eb45bac520489bdbc0917ef" ], "author": { "name": "Ben Hutchings", "email": "bhutchings@solarflare.com", "time": "Tue Aug 17 02:31:15 2010 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Aug 17 02:31:15 2010 -0700" }, "message": "ethtool: Provide a default implementation of ethtool_ops::get_drvinfo\n\nThe driver name and bus address for a net_device can normally be found\nthrough the driver model now. Instead of requiring drivers to provide\nthis information redundantly through the ethtool_ops::get_drvinfo\noperation, use the driver model to do so if the driver does not define\nthe operation. Since ETHTOOL_GDRVINFO no longer requires the driver\nto implement any operations, do not require net_device::ethtool_ops to\nbe set either.\n\nRemove implementations of get_drvinfo and ethtool_ops that provide\nonly this information.\n\nSigned-off-by: Ben Hutchings \u003cbhutchings@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e78483c5aeb0d7fbb0e365802145f1045e62957e", "tree": "cc1c4733486f562b26ab812ce38ff8026da70007", "parents": [ "9fe6206f400646a2322096b56c59891d530e8d51", "b20d02e37eeabf34b40c3995c2dbb0af53da3c57", "20802224298ce9dfd99a7e26b675fc0c8ae26cac" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Aug 02 09:33:25 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Aug 02 10:09:04 2010 +0200" }, "message": "Merge firewire branches to be released post v2.6.35\n\nConflicts:\n\tdrivers/firewire/core-card.c\n\tdrivers/firewire/core-cdev.c\n\nand forgotten #include \u003clinux/time.h\u003e in drivers/firewire/ohci.c\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "20802224298ce9dfd99a7e26b675fc0c8ae26cac", "tree": "a00ead28ffc6b5032f1d83a22f7503fdc2c1f0ec", "parents": [ "872e330e38806d835bd6c311c93ab998e2fb9058" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Aug 01 12:23:14 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Aug 02 08:59:52 2010 +0200" }, "message": "firewire: core: add forgotten dummy driver methods, remove unused ones\n\nThere is an at least theoretic race condition in which .start_iso etc.\ncould still be called between when the dummy driver is bound to the card\nand when the children devices are being shut down. Add dummy_start_iso\nand friends.\n\nOn the other hand, .enable, .set_config_rom, .read_csr, write_csr do not\nneed to be implemented by the dummy driver, as commented.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "872e330e38806d835bd6c311c93ab998e2fb9058", "tree": "92497ce79b1157761b1aebdb63b8d74f68d42c15", "parents": [ "ae2a97661482c1d0f1aa41b837da95054d0e9a1b" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 29 18:19:22 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 29 23:09:18 2010 +0200" }, "message": "firewire: add isochronous multichannel reception\n\nThis adds the DMA context programming and userspace ABI for multichannel\nreception, i.e. for listening on multiple channel numbers by means of a\nsingle DMA context.\n\nThe use case is reception of more streams than there are IR DMA units\noffered by the link layer. This is already implemented by the older\nohci1394 + ieee1394 + raw1394 stack. And as discussed recently on\nlinux1394-devel, this feature is occasionally used in practice.\n\nThe big drawbacks of this mode are that buffer layout and interrupt\ngeneration necessarily differ from single-channel reception: Headers\nand trailers are not stripped from packets, packets are not aligned with\nbuffer chunks, interrupts are per buffer chunk, not per packet.\n\nThese drawbacks also cause a rather hefty code footprint to support this\nrarely used OHCI-1394 feature. (367 lines added, among them 94 lines of\nadded userspace ABI documentation.)\n\nThis implementation enforces that a multichannel reception context may\nonly listen to channels to which no single-channel context on the same\nlink layer is presently listening to. OHCI-1394 would allow to overlay\nsingle-channel contexts by the multi-channel context, but this would be\na departure from the present first-come-first-served policy of IR\ncontext creation.\n\nThe implementation is heavily based on an earlier one by Jay Fenlason.\nThanks Jay.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "ae2a97661482c1d0f1aa41b837da95054d0e9a1b", "tree": "d80b34cc5895c18af5e092ff7c56c1cefb0e1983", "parents": [ "69e61d0c07fa28a05f699723a88d49e0014019b6" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 29 09:31:56 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 29 23:09:18 2010 +0200" }, "message": "firewire: core: small clarifications in core-cdev\n\nMake a note on the seemingly unused linux/sched.h.\nRename an irritatingly named variable.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "69e61d0c07fa28a05f699723a88d49e0014019b6", "tree": "dc17f40f9765506b511b31a2a4afddc2f2c938c4", "parents": [ "e5b06c077c592e7e1623641520787a3da7b7c6bf" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jul 28 23:49:45 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 29 23:06:25 2010 +0200" }, "message": "firewire: core: remove unused code\n\nioctl_create_iso_context enforces ctx-\u003eheader_size \u003e\u003d 4.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "e5b06c077c592e7e1623641520787a3da7b7c6bf", "tree": "92669edec5be2e06ed1b971e611a7683bb33ab75", "parents": [ "071595ebdc66d70219e2d1ce746016f64b2b19e7" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jul 28 15:50:00 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 29 23:06:25 2010 +0200" }, "message": "firewire: ohci: release channel in error path\n\nfirewire-ohci keeps book of which isochronous channels are occupied by\nIR DMA contexts, so that there cannot be more than one context listening\nto a certain channel.\n\nIf IR context creation failed due to an out-of-memory condition, this\nbookkeeping leaked a channel.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "071595ebdc66d70219e2d1ce746016f64b2b19e7", "tree": "3083b0de17a6d0633420ffda1dfa04a0b4b880b2", "parents": [ "8e2b2b46ea4ca5ef790dddf78b360ed736a62d7c" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 13:20:33 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 29 23:06:25 2010 +0200" }, "message": "firewire: ohci: use memory barriers to order descriptor updates\n\nWhen we append to a DMA program, we need to ensure that the order in\nwhich initialization of the new descriptors and update of the\nbranch_address of the old tail descriptor, as seen by the PCI device,\nhappen as intended.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "9f6d3c4b76314c40c866a935d78c80fd284768bd", "tree": "ccb994e8ebe7a36b8bed804c79ce9f117455f041", "parents": [ "aed69d2b79bb5af008526998e466da6d0eac7ae5" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:58:05 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:11 2010 +0200" }, "message": "tools/firewire: add userspace front-end of nosy\n\nThis adds nosy-dump, the userspace part of nosy, the IEEE 1394 traffic\nsniffer for Texas Instruments PCILynx/ PCILynx2 based cards. Author is\nKristian Høgsberg.\n\nThe files added here are taken from\ngit://anongit.freedesktop.org/~krh/nosy commit ee29be97 (2009-11-10)\nwith the following changes by Stefan Richter:\n - Parts pertaining to the kernel module removed from Makefile.\n - dist target removed from the Makefile.\n - Mentioned nosy-dump in the Kconfig help to nosy\u0027s kernel component.\n - Add copyright notice to nosy-dump.c. This is a duplicate of the\n respective notice in the kernel component nosy.c except for a time\n span of 2002 - 2006, according to Kristian\u0027s git log.\n\n\"git shortlog decode-fcp.c list.h nosy-dump.[ch]\" from nosy\u0027s git\nrepository:\n\nJonathan Woithe (1):\n Save logs on Ctrl-C\n\nKristian Høgsberg (11):\n Pull over nosy from mercurial repo.\n Remove some fields from default view, add logging feature.\n Use infinite time out for poll(), mark more detail fields.\n Fix byte ordering macro.\n Add decoding of iso data and lock packets.\n Add flag to indicate data length field.\n Add cycle start packet decoding, add --iso and --cycle-start flags.\n Distinguish between phy-packets and 0-length iso data.\n Fix transaction and stats view.\n Add simple AV/C decoder.\n Don\u0027t break down on big payloads.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\nAcked-by: Kristian Høgsberg \u003ckrh@bitplanet.net\u003e\n" }, { "commit": "7429b17d30a19fd52a0c07de9d3959746d321e15", "tree": "2dc1e07547893ed804fd06b84eb24a86a38cc571", "parents": [ "fd8c8d46ca9402c15383d2cf0bc3ee7740de3b62" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:11 2010 +0200" }, "message": "firewire: nosy: use generic printk macros\n\nReplace home-grown printk wrapper macros by ones from kernel.h and\ndevice.h.\n\nAlso raise the log level in set_phy_reg() from debug to error because\nthese are really error conditions. Could even be WARN_ON. Lower the\nlog level in the device probe and driver shutdown from notice to info.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "fd8c8d46ca9402c15383d2cf0bc3ee7740de3b62", "tree": "66a33db3ea837f2fb6c66e077a9b5fae7009c3ba", "parents": [ "c89db7b8bc88d8288dcfbe7a885b950d2560d564" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:11 2010 +0200" }, "message": "firewire: nosy: endianess fixes and annotations\n\n1.) The DMA programs (struct pcl) are PCI-endian \u003d little endian data\n(except for the 3rd quadlet in a PCL which the controller does not\ntouch). Annotate them as such.\n\nFix all accesses of the PCL to work with big endian CPUs also. Not\nactually tested, I only have a little endian PC to test with. This\nincludes replacement of a bitfield struct pcl_status by open-coded\nshift and mask operations.\n\n2.) The two __attribute__ ((packed)) at struct pcl are not really\nrequired since it consists of u32/__le32 only, i.e. there will be no\npadding with or without the attribute.\n\n3.) The received IEEE 1394 data are byteswapped by the controller from\nIEEE 1394 endian \u003d big endian to PCI endian \u003d little endian because the\nPCL_BIGENDIAN control bit is set. Therefore annotate the DMA buffer as\na __le32 array.\n\nFix the one access of the DMA buffer (the check of the transaction code\nof link packets) to work with big endian CPUs. Also fix the two\naccesses of the client bounce buffer (the reading of packet length).\n\n4.) Add a comment to the userspace ABI header that all of the data gets\nout as little endian data, except for the timestamp which is CPU endian.\n(We could make it little endian too, but why? Vice versa, an ioctl\ncould be added to dump packet data in big endian byte order...)\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "c89db7b8bc88d8288dcfbe7a885b950d2560d564", "tree": "b6588c1c885f6219a55aff794681e39062408ad4", "parents": [ "424d66cedae8bebb00fdb917fc8430f7b8a655cf" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:11 2010 +0200" }, "message": "firewire: nosy: annotate __user pointers and __iomem pointers\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "424d66cedae8bebb00fdb917fc8430f7b8a655cf", "tree": "cd232df29974be404978d47a2257c9422272d304", "parents": [ "b6d9c125e6610591c04ca9045f641e35ce1a9226" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:11 2010 +0200" }, "message": "firewire: nosy: fix device shutdown with active client\n\nFix race between nosy_open() and remove_card() by replacing the\nunprotected array of card pointers by a mutex-protected list of cards.\n\nMake card instances reference-counted and let each client hold a\nreference.\n\nNotify clients about card removal via POLLHUP in poll()\u0027s events\nbitmap; also let read() fail with errno\u003dENODEV if the card was removed\nand everything in the buffer was read.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "b6d9c125e6610591c04ca9045f641e35ce1a9226", "tree": "416672c79a3ee4b0764561ba5ebcf9ed3ba4d5c7", "parents": [ "165476671f731b4c3d6cf401d0e1886f4a4f4a8e" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: nosy: handle errors in device probe\n\nand add a missing pci_disable_device() to device shutdown.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "165476671f731b4c3d6cf401d0e1886f4a4f4a8e", "tree": "59bba3913c7bb3a023394bce86bb0ea79d95f1ed", "parents": [ "55e77c06c6017a70630cf599770369b8ba07c841" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: nosy: fix IRQ handler for card ejection\n\nUntested, I don\u0027t have a PCILynx CardBus card.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "55e77c06c6017a70630cf599770369b8ba07c841", "tree": "5fe5940bd0e0326afad0898b521490b53c894adf", "parents": [ "685c3f80b6d88478a6428676f9daab59faf3cd4b" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: nosy: unroll some simple functions\n\nnosy_start/stop_snoop() and nosy_add/remove_client() are simple enough\nto be inlined into their callers.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "685c3f80b6d88478a6428676f9daab59faf3cd4b", "tree": "3f1f93927c0241ba12d887bcf7b90fc52d760884", "parents": [ "a2d39db9dec0e7e403f54c9cf98b7dbc82b4c44a" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: nosy: use flagless variants of spinlock accessors\n\nnosy_start/stop_snoop() are always only called by the ioctl method, i.e.\nwith IRQs enabled. packet_handler() and bus_reset_handler() are always\nonly called by the IRQ handler. Hence neither one needs to track IRQ\nflags.\n\nTo underline the call context of packet_handler() and\nbus_reset_handler(), rename these functions to *_irq_handler().\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "a2d39db9dec0e7e403f54c9cf98b7dbc82b4c44a", "tree": "86e4e8447f83cefa4ae4809051fffad04b8229ef", "parents": [ "c7b2a99c66e7b40d8843a70f2981e375eeedf062" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: nosy: fix list corruption by NOSY_IOC_STOP\n\nnosy_stop_snoop() would blow up the second time it was called without\nnosy_start_snoop() in between.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "c7b2a99c66e7b40d8843a70f2981e375eeedf062", "tree": "56039d14506b685e009f0b501264cba08d3e5484", "parents": [ "b5e47729043c9224b21ab3dc7c63e8a38dbb4923" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 22 11:56:38 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: nosy: convert to unlocked ioctl\n\nThe required serialization of NOSY_IOC_START and NOSY_IOC_STOP is\nalready provided by the client_list_lock.\n\nNOSY_IOC_FILTER does not really require serialization since accesses\nto tcode_mask are atomic on any sane CPU architecture. Nevertheless,\nmake it explicit that we want this to be atomic by means of\nclient_list_lock (which also surrounds the other tcode_mask access in\nthe IRQ handler). While we are at it, change the type of tcode_mask to\nu32 for consistency with the user API.\n\nNOSY_IOC_GET_STATS does not require serialization against itself. But\nthere is a bug here regarding concurrent updates of the two counters\nby the IRQ handler. Fix it by taking the client_list_lock in this ioctl\ntoo.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "b5e47729043c9224b21ab3dc7c63e8a38dbb4923", "tree": "2acf6aa0f441c3a6b8812c8c5eb3b97a74f98056", "parents": [ "286468210d83ce0ca1e37e346ed9f4457a161650" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 10:28:30 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: nosy: misc cleanups\n\nExtend copyright note to 2007, c.f. Kristian\u0027s git log.\n\nIncludes:\n - replace some \u003casm/*.h\u003e by \u003clinux/*.h\u003e\n - add required indirectly included \u003clinux/spinlock.h\u003e\n - order alphabetically\n\nCoding style related changes:\n - change to utf8\n - normalize whitespace\n - normalize comment style\n - remove usages of __FUNCTION__\n - remove an unnecessary cast from void *\n\nConst and static declarations:\n - driver_name is not const in pci_driver.name, drop const qualifier\n - driver_name can be taken from KBUILD_MODNAME\n - the global variable minors[] can and should be static\n - constify struct file_operations instance\n\nData types:\n - Remove unused struct member struct packet.code. struct packet is\n only used for driver-internal bookkeeping; it does not appear on the\n wire or in DMA programs or the userspace ABI. Hence the unused\n member .code can be removed without worries.\n\nPreprocessor macros:\n - unroll a preprocessor macro that containd a return\n - use list_for_each_entry\n\nPrintk:\n - add missing terminating \\n in some format strings\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "286468210d83ce0ca1e37e346ed9f4457a161650", "tree": "e445a09a6a074e3ae65479e417d41d4d1c41f571", "parents": [ "e40152ee1e1c7a63f4777791863215e3faa37a86" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 10:26:33 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 27 11:04:10 2010 +0200" }, "message": "firewire: new driver: nosy - IEEE 1394 traffic sniffer\n\nThis adds the traffic sniffer driver for Texas Instruments PCILynx/\nPCILynx2 based cards. The use cases for nosy are analysis of\nnonstandard protocols and as an aid in development of drivers,\napplications, or firmwares.\n\nAuthor of the driver is Kristian Høgsberg. Known contributers are\nJody McIntyre and Jonathan Woithe.\n\nNosy programs PCILynx chips to operate in promiscuous mode, which is a\nfeature that is not found in OHCI-1394 controllers. Hence, only special\nhardware as mentioned in the Kconfig help text is suitable for nosy.\n\nThis is only the kernelspace part of nosy. There is a userspace\ninterface to it, called nosy-dump, proposed to be added into the tools/\nsubdirectory of the kernel sources in a subsequent change. Kernelspace\nand userspave component of nosy communicate via a \u0027misc\u0027 character\ndevice file called /dev/nosy with a simple ioctl() and read() based\nprotocol, as described by nosy-user.h.\n\nThe files added here are taken from\ngit://anongit.freedesktop.org/~krh/nosy commit ee29be97 (2009-11-10)\nwith the following changes by Stefan Richter:\n - Kconfig and Makefile hunks are written from scratch.\n - Commented out version printk in nosy.c.\n - Included missing \u003clinux/sched.h\u003e, reported by Stephen Rothwell.\n\n\"git shortlog nosy{-user.h,.c,.h}\" from nosy\u0027s git repository:\n\nJonathan Woithe (2):\n Nosy updates for recent kernels\n Fix uninitialised memory (needed for 2.6.31 kernel)\n\nKristian Høgsberg (5):\n Pull over nosy from mercurial repo.\n Use a misc device instead.\n Add simple AV/C decoder.\n Don\u0027t break down on big payloads.\n Set parent device for misc device.\n\nAs a low-level IEEE 1394 driver, its files are placed into\ndrivers/firewire/ although nosy is not part of the firewire driver\nstack.\n\nI am aware of the following literature from Texas Instruments about\nPCILynx programming:\n SCPA020A - PCILynx 1394 to PCI Bus Interface TSB12LV21BPGF\n Functional Specification\n SLLA023 - Initialization and Asynchronous Programming of the\n TSB12LV21A 1394 Device\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\nAcked-by: Kristian Høgsberg \u003ckrh@bitplanet.net\u003e\n" }, { "commit": "8e2b2b46ea4ca5ef790dddf78b360ed736a62d7c", "tree": "35fb2be1225c5a4733d4be8c03bc5725f79b43c2", "parents": [ "0c9ae701ae1caf657326db22d61074b40a747c9d" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:05:39 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:36:28 2010 +0200" }, "message": "firewire: cdev: improve FW_CDEV_IOC_ALLOCATE\n\nIn both the ieee1394 stack and the firewire stack, the core treats\nkernelspace drivers better than userspace drivers when it comes to\nCSR address range allocation: The former may request a register to be\nplaced automatically at a free spot anywhere inside a specified address\nrange. The latter may only request a register at a fixed offset.\n\nHence, userspace drivers which do not require a fixed offset potentially\nneed to implement a retry loop with incremented offset in each retry\nuntil the kernel does not fail allocation with EBUSY. This awkward\nprocedure is not fundamentally necessary as the core already provides a\nsuperior allocation API to kernelspace drivers.\n\nTherefore change the ioctl() ABI by addition of a region_end member in\nthe existing struct fw_cdev_allocate. Userspace and kernelspace APIs\nwork the same way now.\n\nThere is a small cost to pay by clients though: If client source code\nis required to compile with older kernel headers too, then any use of\nthe new member fw_cdev_allocate.region_end needs to be enclosed by\n#ifdef/#endif directives. However, any client program that seriously\nwants to use address range allocations will require a kernel of cdev ABI\nversion \u003e\u003d 4 at runtime and a linux/firewire-cdev.h header of \u003e\u003d 4\nanyway. This is because v4 brings FW_CDEV_EVENT_REQUEST2. The only\nclient program in which build-time compatibility with struct\nfw_cdev_allocate as found in older kernel headers makes sense is\nlibraw1394.\n\n(libraw1394 uses the older broken FW_CDEV_EVENT_REQUEST to implement a\nmakeshift, incorrect transaction responder that does at least work\nsomewhat in many simple scenarios, relying on guesswork by libraw1394\nand by libraw1394 based applications. Plus, address range allocation\nand transaction responder is only one of many features that libraw1394\nneeds to provide, and these other features need to work with kernel and\nkernel-headers as old as possible. Any new linux/firewire-cdev.h based\nclient that implements a transaction responder should never attempt to\ndo it like libraw1394; instead it should make a header and kernel of v4\nor later a hard requirement.)\n\nWhile we are at it, update the struct fw_cdev_allocate documentation to\nbetter reflect the recent fw_cdev_event_request2 ABI addition.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "0c9ae701ae1caf657326db22d61074b40a747c9d", "tree": "97eccc9b9941e71c471b5b3f32450c89f476093f", "parents": [ "cc550216ae9a2993ef3973464714dc1a39ab1f86" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:02:54 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:36:28 2010 +0200" }, "message": "firewire: core: fix upper bound of possible CSR allocations\n\nregion-\u003eend is defined as an upper bound of the requested address range,\nexclusive --- i.e. as an address outside of the range in which the\nrequested CSR is to be placed.\n\nHence 0x0001,0000,0000,0000 is the biggest valid region-\u003eend, not\n0x0000,ffff,ffff,fffc like the current check asserted.\n\nFor simplicity, the fix drops the region-\u003eend \u0026 3 test because there is\nno actual problem with these bits set in region-\u003eend. The allocated\naddress range will be quadlet aligned and of a size of multiple quadlets\ndue to the checks for region-\u003estart \u0026 3 and handler-\u003elength \u0026 3 alone.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "cc550216ae9a2993ef3973464714dc1a39ab1f86", "tree": "62ffde836c83fe44b7a9edc01d00bcb9ad4f4ad6", "parents": [ "bf54e1462b9192fdef7ea9e2bc44fdc16a4b87bc" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jul 18 13:00:50 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:36:28 2010 +0200" }, "message": "firewire: cdev: add PHY pinging\n\nThis extends the FW_CDEV_IOC_SEND_PHY_PACKET ioctl() for /dev/fw* to be\nuseful for ping time measurements. One application for it would be gap\ncount optimization in userspace that is based on ping times rather than\nhop count. (The latter is implemented in firewire-core itself but is\nnot applicable to beta PHYs that act as repeater.)\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "bf54e1462b9192fdef7ea9e2bc44fdc16a4b87bc", "tree": "31ec8e4e13b76d22b7bf9f93ea620e88911fe416", "parents": [ "850bb6f23b93c04ce1e4509a87fa607dc17d97c1" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 16 22:25:51 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:36:28 2010 +0200" }, "message": "firewire: cdev: add PHY packet reception\n\nAdd an FW_CDEV_IOC_RECEIVE_PHY_PACKETS ioctl() and\nFW_CDEV_EVENT_PHY_PACKET_RECEIVED poll()/read() event for /dev/fw*.\nThis can be used to get information from remote PHYs by remote access\nPHY packets.\n\nThis is also the 2nd half of the functionality (the receive part) to\nsupport a userspace implementation of a VersaPHY transaction layer.\n\nSafety considerations:\n\n - PHY packets are generally broadcasts, hence some kind of elevated\n privileges should be required of a process to be able to listen in\n on PHY packets. This implementation assumes that a process that is\n allowed to open the /dev/fw* of a local node does have this\n privilege.\n\n There was an inconclusive discussion about introducing POSIX\n capabilities as a means to check for user privileges for these\n kinds of operations.\n\nOther limitations:\n\n - PHY packet reception may be switched on by ioctl() but cannot be\n switched off again. It would be trivial to provide an off switch,\n but this is not worth the code. The client should simply close()\n the fd then, or just ignore further events.\n\n - For sake of simplicity of API and kernel-side implementation, no\n filter per packet content is provided.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "850bb6f23b93c04ce1e4509a87fa607dc17d97c1", "tree": "4ec66cea35e15ff095799bae4aec7c1071d4faa4", "parents": [ "b9dc61cf404165fb77e80c853e9fec9af258f9ce" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 16 22:25:14 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:36:28 2010 +0200" }, "message": "firewire: cdev: add PHY packet transmission\n\nAdd an FW_CDEV_IOC_SEND_PHY_PACKET ioctl() for /dev/fw* which can be\nused to implement bus management related functionality in userspace.\n\nThis is also half of the functionality (the transmit part) that is\nneeded to support a userspace implementation of a VersaPHY transaction\nlayer.\n\nSafety considerations:\n\n - PHY packets are generally broadcasts and may have interesting\n effects on PHYs and the bus, e.g. make asynchronous arbitration\n impossible due to too low gap count. Hence some kind of elevated\n privileges should be required of a process to be able to send\n PHY packets. This implementation assumes that a process that is\n allowed to open the /dev/fw* of a local node does have this\n privilege.\n\n There was an inconclusive discussion about introducing POSIX\n capabilities as a means to check for user privileges for these\n kinds of operations.\n\n - The kernel does not check integrity of the supplied packet data.\n That would be far too much code, considering the many kinds of\n PHY packets. A process which got the privilege to send these\n packets is trusted to do it correctly.\n\nJust like with the other \"send packet\" ioctls, a non-blocking API is\nchosen; i.e. the ioctl may return even before AT DMA started. After\ntransmission, an event for poll()/read() is enqueued. Most users are\ngoing to need a blocking API, but a blocking userspace wrapper is easy\nto implement, and the second of the two existing libraw1394 calls\nraw1394_phy_packet_write() and raw1394_start_phy_packet_write() can be\nbetter supported that way.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "b9dc61cf404165fb77e80c853e9fec9af258f9ce", "tree": "7a8f7741b9482218332fa68bfcfcd54583434809", "parents": [ "d505e6e87127d4dbdaa5d91561eed810c180ca23" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 16 22:24:29 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:36:28 2010 +0200" }, "message": "firewire: core: use C99 initializer in array of ioctl handlers\n\nto make the correspondence of ioctl numbers and handlers more obvious.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "18d0cdfd1a4cc9028c0ef80f94538b31541f8fe5", "tree": "c28db7537da311e59e871aba7f6f0edaa46535b9", "parents": [ "80792d182e43bee89fce509e64fdea27e600530f" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jul 18 12:44:01 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Fri Jul 23 13:36:27 2010 +0200" }, "message": "firewire: normalize status values in packet callbacks\n\ncore-transaction.c transmit_complete_callback() and close_transaction()\nexpect packet callback status to be an ACK or RCODE, and ACKs get\ntranslated to RCODEs for transaction callbacks.\n\nAn old comment on the packet callback API (been there from the initial\nsubmission of the stack) and the dummy_driver implementation of\nsend_request/send_response deviated from this as they also included\n-ERRNO in the range of status values.\n\nLet\u0027s narrow status values down to ACK and RCODE to prevent surprises.\nRCODE_CANCELLED is chosen as the dummy_driver\u0027s RCODE as its meaning of\n\"transaction timed out\" comes closest to what happens when a transaction\ncoincides with card removal.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "02d37bed188c500ee7afb0a2dc6b65a80704c58e", "tree": "a019891672a1505e35eb15fa2621caffecff2c80", "parents": [ "8b4f70ba4967cae90d128857af1382026a24230a" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 08 16:09:06 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 13 09:58:27 2010 +0200" }, "message": "firewire: core: integrate software-forced bus resets with bus management\n\nBus resets which are triggered\n - by the kernel drivers after updates of the local nodes\u0027 config ROM,\n - by userspace software via ioctl\nshall be deferred until after \u003e\u003d2 seconds after the last bus reset.\n\nIf multiple modifications of the local nodes\u0027 config ROM happen in a row,\nonly a single bus reset should happen after them.\n\nWhen the local node\u0027s link goes from inactive to active or vice versa,\nand at the two occasions of bus resets mentioned above --- and if the\ncurrent gap count differs from 63 --- the bus reset should be preceded\nby a PHY configuration packet that reaffirms the gap count. Otherwise a\nbus manager would have to reset the bus again right after that.\n\nThis is necessary to promote bus stability, e.g. leave grace periods for\nallocations and reallocations of isochronous channels and bandwidth,\nSBP-2 reconnections etc.; see IEEE 1394 clause 8.2.1.\n\nThis change implements all of the above by moving bus reset initiation\ninto a delayed work (except for bus resets which are triggered by the\nbus manager workqueue job and are performed there immediately). It\ncomes with a necessary addition to the card driver methods that allows\nto get the current gap count from PHY registers.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "8b4f70ba4967cae90d128857af1382026a24230a", "tree": "a327de21c9c5758bef562df9cb21ac90980b1c70", "parents": [ "eb5b35a560510efc6bb62f05c3c82e9596cdfafe" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jul 07 15:36:07 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 13 09:47:47 2010 +0200" }, "message": "firewire: cdev: fix fw_cdev_event_bus_reset emission after local config ROM changes\n\nWhen a descriptor was added or removed to the local node\u0027s config ROM,\nuserspace clients which had a local node\u0027s /dev/fw* open did not receive\nany fw_cdev_event_bus_reset for poll()/read() consumption.\n\nThe cause was that the core-device.c facility which re-reads the config\nROM of the bus reset initiator node missed to call the fw_device update\nfunction. The fw_units are destroyed and newly added, but their parent\nstays and needs to be updated.\n\nReported-by: Jay Fenlason \u003cfenlason@redhat.com\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "eb5b35a560510efc6bb62f05c3c82e9596cdfafe", "tree": "59d610352e8e0e60de0ce958149fb08dddf6ce82", "parents": [ "656b7afd40a9f2b0d6cf8ef1972681961b428558" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jul 07 14:13:14 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 13 09:47:47 2010 +0200" }, "message": "firewire: core: ensure some userspace API constants match corresponding kernel API constants\n\nThe FW_ISO_ constants of the in-kernel API of firewire-core and\nFW_CDEV_ISO_ constants of the userspace API of firewire-core have\nnothing to do with each other --- except that the core-cdev.c\nimplementation relies on them having the same values.\n\nHence put some compile-time assertions into core-cdev.c. It\u0027s lame but\nI prefer it over including the userspace API header into the kernelspace\nAPI header and defining kernelspace API constants from userspace API\nconstants. Nor do I want to expose the kernelspace constants in one of\nthe two firewire headers that are exported to userland since this only\nconcerns the core-cdev.c implementation.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "656b7afd40a9f2b0d6cf8ef1972681961b428558", "tree": "959bc229119224651799c2c1fce388a78cae08b8", "parents": [ "a8e93f3dccc066cd6dd1e9db1e35942914fc57d1" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jul 07 13:26:18 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 13 09:47:47 2010 +0200" }, "message": "firewire: core: fix fw_send_request kerneldoc comment\n\nThe present inline documentation of the fw_send_request() in-kernel API\nrefers to userland code that is not applicable to kernel drivers at all.\n\nReported-by: Ben Gamari \u003cbgamari.foss@gmail.com\u003e\n\nWhile we are at fixing the whole documentation of fw_send_request(),\nalso improve the rest of firewire-core\u0027s kerneldoc comments:\n - Add a bit of text concerning fw_run_transaction()\u0027s call parameters.\n - Append () to function names and tab-align parameter descriptions as\n suggested by the example in Documentation/kernel-doc-nano-HOWTO.txt.\n - Remove kerneldoc markers from comments on static functions.\n - Remove outdated parameter descriptions at build_tree().\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "a8e93f3dccc066cd6dd1e9db1e35942914fc57d1", "tree": "9165f872029ef76e99fd40c0afb6d8c896f8cabc", "parents": [ "250b2b6dd421c9f8844a867d2ac06e0661e0ad93" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Wed Jul 07 14:37:30 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jul 13 09:47:47 2010 +0200" }, "message": "firewire: cdev: check write quadlet request length to avoid buffer overflow\n\nCheck that the data length of a write quadlet request actually is large\nenough for a quadlet. Otherwise, fw_fill_request could access the four\nbytes after the end of the outbound_transaction_event structure.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n\nModification of Clemens\u0027 change: Consolidate the check into\ninit_request() which is used by the affected ioctl_send_request() and\nioctl_send_broadcast_request() and the unaffected\nioctl_send_stream_packet(), to save a few lines of code.\n\nNote, since struct outbound_transaction_event *e is slab-allocated, such\nan out-of-bounds access won\u0027t hit unallocated memory but may result in a\n(virtually impossible to exploit) information disclosure.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "250b2b6dd421c9f8844a867d2ac06e0661e0ad93", "tree": "b86719d2c442676a52c6307471c9e278b1bd21f5", "parents": [ "ae948011071c12ff6a328348859c717ea885ed40" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Jun 21 23:24:35 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 08 16:52:02 2010 +0200" }, "message": "firewire: cdev: fix fw_cdev_event_bus_reset.bm_node_id\n\nFix an obscure ABI feature that is a bit of a hassle to implement.\nHowever, somebody put it into the ABI, so let\u0027s fill in a sensible\nvalue there.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "ae948011071c12ff6a328348859c717ea885ed40", "tree": "7ca46403da2d288ca64e0d858dbe7cfc07e34a6d", "parents": [ "3b2b65d68fc87b02ac393a031a4ebb3de84a8218" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Mon Jun 21 23:23:52 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Thu Jul 08 16:45:54 2010 +0200" }, "message": "firewire: core: no need to track irq flags in bm_work\n\nThis is a workqueue job and always entered with IRQs enabled.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "e205597d188a9ea69ce43f740a14f07b3f5b996a", "tree": "9b894920db99c784c9abbe95269b972d0a1fc5c7", "parents": [ "c82f91f2663e79b150afd896ec72e798ba4e243d" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 22:53:55 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 23:11:56 2010 +0200" }, "message": "firewire: cdev: fix ABI for FCP and address range mapping, add fw_cdev_event_request2\n\nThe problem:\n\nA target-like userspace driver, e.g. AV/C target or SBP-2/3 target,\nneeds to be able to act as responder and requester. In the latter role,\nit needs to send requests to nods from which it received requests. This\nis currently impossible because fw_cdev_event_request lacks information\nabout sender node ID.\nReported-by: Jay Fenlason \u003cfenlason@redhat.com\u003e\n\nLibffado + libraw1394 + firewire-core is currently unable to drive two\nor more audio devices on the same bus.\nReported-by: Arnold Krille \u003carnold@arnoldarts.de\u003e\n\nThis is because libffado requires destination node ID of FCP requests\nand sender node ID of FCP responses to match. It even prohibits\nlibffado from working with a bus on which libraw1394 opens a /dev/fw* as\ndefault ioctl device that does not correspond with the audio device.\nThis is because libraw1394 does not receive the sender node ID from the\nkernel.\n\nMoreover, fw_cdev_event_request makes it impossible to tell unicast and\nbroadcast write requests apart.\n\nThe fix:\n\nAdd a replacement of struct fw_cdev_event_request request, boringly\ncalled struct fw_cdev_event_request2. The new event will be sent to a\nuserspace client instead of the old one if the client claims\ncompatibility with \u003clinux/firewire-cdev.h\u003e ABI version 4 or later.\n\nlibraw1394 needs to be extended to make use of the new event, in order\nto properly support libffado and other FCP or address range mapping\nusers who require correct sender node IDs.\n\nFurther notes:\n\nWhile we are at it, change back the range of possible values of\nfw_cdev_event_request.tcode to 0x0...0xb like in ABI version \u003c\u003d 3.\nThe preceding change \"firewire: expose extended tcode of incoming lock\nrequests to (userspace) drivers\" expanded it to 0x0...0x17 which could\ncatch sloppily coded clients by surprise. The extended range of codes\nis only used in the new fw_cdev_event_request2.tcode.\n\nJay and I also suggested an alternative approach to fix the ABI for\nincoming requests: Add an FW_CDEV_IOC_GET_REQUEST_INFO ioctl which can\nbe called after reception of an fw_cdev_event_request, before issuing of\nthe closing FW_CDEV_IOC_SEND_RESPONSE ioctl. The new ioctl would reveal\nthe vital information about a request that fw_cdev_event_request lacks.\nJay showed an implementation of this approach.\n\nThe former event approach adds 27 LOC of rather trivial code to\ncore-cdev.c, the ioctl approach 34 LOC, some of which is nontrivial.\nThe ioctl approach would certainly also add more LOC to userspace\nprograms which require the expanded information on inbound requests.\nThis approach is probably only on the lighter-weight side in case of\nclients that want to be compatible with kernels that lack the new\ncapability, like libraw1394. However, the code to be added to such\nlibraw1394-like clients in case of the event approach is a straight-\nforward additional switch () case in its event handler.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "c82f91f2663e79b150afd896ec72e798ba4e243d", "tree": "12555e4299cb8cb7e958ff7576b6a5be2b781e0d", "parents": [ "604f45167824e18ad5766e51ecf1d4d65f15118d" ], "author": { "name": "Jay Fenlason", "email": "fenlason@redhat.com", "time": "Tue May 18 10:57:33 2010 -0400" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 23:11:56 2010 +0200" }, "message": "firewire: expose extended tcode of incoming lock requests to (userspace) drivers\n\nWhen a remote device does a LOCK_REQUEST, the core does not pass\nthe extended tcode to userspace. This patch makes it use the\njuju-specific tcodes listed in firewire-constants.h for incoming\nrequests.\n\nSigned-off-by: Jay Fenlason \u003cfenlason@redhat.com\u003e\n\nThis matches how tcode in the API for outbound requests is treated.\nAffects kernelspace and userspace drivers alike, but at the moment there\nare no kernespace drivers that receive lock requests.\n\nSplit out from a combo patch, slightly reordered, changelog reworded.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "604f45167824e18ad5766e51ecf1d4d65f15118d", "tree": "40deeca88e1788e98d54d82f429a45ce276c4135", "parents": [ "0244f57302f7e8bebd2f1ab58767eac2e9f678a6" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 22:52:55 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 23:11:56 2010 +0200" }, "message": "firewire: cdev: freeze FW_CDEV_VERSION due to libraw1394 bug\n\nlibraw1394 v2.0.0...v2.0.5 takes FW_CDEV_VERSION from an externally\ninstalled header file and uses it to declare its own implementation\nlevel in FW_CDEV_IOC_GET_INFO. This is wrong; it should set the real\nversion for which it was actually written.\n\nIf we add features to the kernel ABI that require the kernel to check\na client\u0027s implementation level, we can not trust the client version if\nit was set from FW_CDEV_VERSION.\n\nHence freeze FW_CDEV_VERSION at the current value (no damage has been\ndone yet), clearly document FW_CDEV_VERSION as a dummy version and what\nclients are expected to do with fw_cdev_get_info.version, and use a new\ndefined constant (which is not placed into the exported header file) as\nkernel implementation level.\n\nNote, in order to check in client program source code which features are\npresent in an externally installed linux/firewire-cdev.h, use\npreprocessor directives like\n #ifdef FW_CDEV_IOC_ALLOCATE_ISO_RESOURCE\nor\n #ifdef FW_CDEV_EVENT_ISO_RESOURCE_ALLOCATED\ninstead of a check of FW_CDEV_VERSION.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "0244f57302f7e8bebd2f1ab58767eac2e9f678a6", "tree": "b9ea719d24b41f9df661e8de1dab36bfc3c58b11", "parents": [ "08bd34c98d631fe85744d4c920c80f48a1d95f54" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 22:52:27 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 23:11:56 2010 +0200" }, "message": "firewire: cdev: count references of cards during inbound transactions\n\nIf a request comes in to an address range managed by a userspace driver\ni.e. \u003clinux/firewire-cdev.h\u003e client, the card instance of request and\nresponse may differ from the card instance of the client device.\nTherefore we need to take a reference of the card until the response was\nsent.\n\nI thought about putting the reference counting into core-transaction.c,\nbut the various high-level drivers besides cdev clients (firewire-net,\nfirewire-sbp2, firedtv) use the card pointer in their fw_address_handler\naddress_callback method only to look up devices of which they already\nhold the necessary references. So this seems to be a specific\nfirewire-cdev issue which is better addressed locally.\n\nWe do not need the reference\n - in case of FCP_REQUEST or FCP_RESPONSE requests because then the\n firewire-core will send the split transaction response for us\n already in the context of the request handler,\n - if it is the same card as the client device\u0027s because we hold a\n card reference indirectly via teh client-\u003edevice reference.\nTo keep things simple, we take the reference nevertheless.\n\nJay Fenlason wrote:\n\u003e there\u0027s no way for the core to tell cdev \"this card is gone,\n\u003e kill any inbound transactions on it\", while cdev holds the transaction\n\u003e open until userspace issues a SEND_RESPONSE ioctl, which may be a very,\n\u003e very long time. But when it does, it calls fw_send_response(), which\n\u003e will dereference the card...\n\u003e\n\u003e So how unhappy are we about userspace potentially holding a fw_card\n\u003e open forever?\n\nWhile termination of inbound transcations at card removal could be\nimplemented, it is IMO not worth the effort. Currently, the effect of\nholding a reference of a card that has been removed is to block the\nprocess that called the pci_remove of the card. This is\n - either a user process ran by root. Root can find and kill processes\n that have /dev/fw* open, if desired.\n - a kernel thread (which one?) in case of hot removal of a PCCard or\n ExpressCard.\nThe latter case could be a problem indeed. firewire-core\u0027s card\nshutdown and card release should probably be improved not to block in\nshutdown, just to defer freeing of memory until release.\n\nThis is not a new problem though; the same already always happens with\nthe client-\u003edevice-\u003ecard without the need of inbound transactions or\nother special conditions involved, other than the client not closing the\nfile.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "08bd34c98d631fe85744d4c920c80f48a1d95f54", "tree": "cebfda304248a705e2b46f9d105b3c22705169f1", "parents": [ "bdfe273ee54b29498851fc8058516037d284270c" ], "author": { "name": "Jay Fenlason", "email": "fenlason@redhat.com", "time": "Tue May 18 14:02:45 2010 -0400" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 23:11:56 2010 +0200" }, "message": "firewire: cdev: fix responses to nodes at different card\n\nMy box has two firewire cards in it: card0 and card1.\nMy application opens /dev/fw0 (card 0) and allocates an address space.\nThe core makes the address space available on both cards.\nAlong comes the remote device, which sends a READ_QUADLET_REQUEST to\ncard1. The request gets passed up to my application, which calls\nioctl_send_response().\n\nioctl_send_response() then calls fw_send_response() with card0,\nbecause that\u0027s the card it\u0027s bound to.\nCard0\u0027s driver drops the response, because it isn\u0027t part of\na transaction that it has outstanding.\n\nSo in core-cdev: handle_request(), we need to stash the\ncard of the inbound request in the struct inbound_transaction_resource and\nuse that card to send the response to.\n\nThe hard part will be refcounting the card correctly\nso it can\u0027t get deallocated while we hold a pointer to it.\n\nHere\u0027s a trivial patch, which does not do the card refcounting, but at\nleast demonstrates what the problem is.\n\nNote that we can\u0027t depend on the fact that the core-cdev:client\nstructure holds a card open, because in this case the card it holds\nopen is not the card the request came in on.\n\n..and there\u0027s no way for the core to tell cdev \"this card is gone,\nkill any inbound transactions on it\", while cdev holds the transaction\nopen until userspace issues a SEND_RESPONSE ioctl, which may be a very,\nvery long time. But when it does, it calls fw_send_response(), which\nwill dereference the card...\n\nSo how unhappy are we about userspace potentially holding a fw_card\nopen forever?\n\nSigned-off-by: Jay Fenlason \u003cfenlason@redhat.com\u003e\n\nReference counting to be addressed in a separate change.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e (whitespace)\n" }, { "commit": "bdfe273ee54b29498851fc8058516037d284270c", "tree": "1e4a94774ae5469ee6e4bc9bd5bb1e43db67c511", "parents": [ "33e553fe2b4a983ef34a57ab1440d8d33397bb12" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Mon Jun 14 11:46:25 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 23:11:56 2010 +0200" }, "message": "firewire: cdev: fix race in iso context creation\n\nProtect the client\u0027s iso context pointer against a race that can happen\nwhen more than one creation call is executed at the same time.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "33e553fe2b4a983ef34a57ab1440d8d33397bb12", "tree": "3a7098a935f0b678a6e6a565619d0d990f9122f3", "parents": [ "56d04cb189f955e5167c27944d61aa57ad69b598" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 22:50:35 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 23:11:55 2010 +0200" }, "message": "firewire: remove an unused function argument\n\nvoid (*fw_address_callback_t)(..., int speed, ...) is the speed that a\nremote node chose to transmit a request to us. In case of split\ntransactions, firewire-core will transmit the response at that speed.\n\nUpper layer drivers on the other hand (firewire-net, -sbp2, firedtv, and\nuserspace drivers) cannot do anything useful with that speed datum,\nexcept log it for debug purposes. But data that is merely potentially\n(not even actually) used for debug purposes does not belong into the API.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "56d04cb189f955e5167c27944d61aa57ad69b598", "tree": "3cea2dce19e0a52c215b2a389f45ed51077a2041", "parents": [ "ae86e81e434072be28ff4e9c1be3cc2562be8749" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jun 08 00:20:10 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun Jun 20 17:06:25 2010 +0200" }, "message": "firewire: core: remove an unnecessary zero initialization\n\nAll of the fields of the iso_interrupt_event instance are overwritten\nright after it was allocated.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "ae86e81e434072be28ff4e9c1be3cc2562be8749", "tree": "f73b99c0fa7d96cc94e2ae6110bcc9d5855a02f7", "parents": [ "5030c807907ae90ad21e9220c1a9d592558deba2" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Tue Jun 15 01:22:45 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 19 13:01:41 2010 +0200" }, "message": "firewire: core: remove unused variable\n\nwhich caused gcc 4.6 to warn about\n variable \u0027destination\u0027 set but not used.\n\nSince the hardware ensures that we receive only response packets with\nproper destination node ID (in a given bus generation), we have no use\nfor destination here in the core as well as in upper layers.\n\n(This is different with request packets. There we pass destination node\nID to upper layers because they may for example need to check whether\nthis was an unicast or broadcast request.)\n\nReported-and-Tested-By: Justin P. Mattock \u003cjustinmattock@gmail.com\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "0fcff4e39323f466a47684d7c8ffa77e1be86c8a", "tree": "cdbe9d239ae59c18a48a34d937da3f658c61e873", "parents": [ "b384cf18873da1ed100662aa7373edf5883a1c24" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 12 20:35:52 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 19 13:01:41 2010 +0200" }, "message": "firewire: rename CSR access driver methods\n\nRather than \"read a Control and Status Registers (CSR) Architecture\nregister\" I prefer to say \"read a Control and Status Register\".\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "b384cf18873da1ed100662aa7373edf5883a1c24", "tree": "2c4f488315855554dd430d0cc1e8969333f26d54", "parents": [ "c8a94ded57e9cc2498d401b2f5c856213a3e19fb" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 12 20:35:21 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 19 13:01:41 2010 +0200" }, "message": "firewire: core: combine some repeated code\n\nAll of these CSRs have the same read/ write/ aynthing-else handling,\nexcept for CSR_PRIORITY_BUDGET which might not be implemented.\n\nThe CSR_CYCLE_TIME read handler implementation accepted 4-byte-sized\nblock write requests before this change but this is just silly; the\nregister is only required to support quadlet read and write requests\nlike the other r/w CSR core and Serial-Bus-dependent registers.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "c8a94ded57e9cc2498d401b2f5c856213a3e19fb", "tree": "8fd6a196ff953270c03700dd682108baf3ee369b", "parents": [ "db3c9cc105ee844f6cd7a1beb9926fb8e9a093ae" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 12 20:34:50 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 19 13:01:41 2010 +0200" }, "message": "firewire: normalize STATE_CLEAR/SET CSR access interface\n\nPush the maintenance of STATE_CLEAR/SET.abdicate down into the card\ndriver. This way, the read/write_csr_reg driver method works uniformly\nacross all CSR offsets.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "db3c9cc105ee844f6cd7a1beb9926fb8e9a093ae", "tree": "5fe981b7be4ffb97a506e23836d6c81a61dc8c91", "parents": [ "e847cc832bab50aad939a0c30414cd986637564d" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 12 20:30:21 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 19 13:01:41 2010 +0200" }, "message": "firewire: replace get_features card driver hook\n\nby feature variables in the fw_card struct. The hook appeared to be an\nunnecessary abstraction in the card driver interface.\n\nCleaner would be to pass those feature flags as arguments to\nfw_card_initialize() or fw_card_add(), but the FairnessControl register\nis in the SCLK domain and may therefore not be accessible while Link\nPower Status is off, i.e. before the card-\u003edriver-\u003eenable call from\nfw_card_add().\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "e847cc832bab50aad939a0c30414cd986637564d", "tree": "35b5d7af3e3f3337b8f7c9c43029618bec3a1f69", "parents": [ "65b2742ac002f554f6ca0c2eab84cc62400eafdf" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 12 20:29:07 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 19 13:01:40 2010 +0200" }, "message": "firewire: drop sizeof expressions from some request size arguments\n\nIn case of fw_card_bm_work()\u0027s lock request, the present sizeof\nexpression is going to be wrong if somebody changes the fw_card\u0027s DMA\nscratch buffer\u0027s size in the future.\n\nIn case of quadlet write requests, sizeof(u32) is just silly; it\u0027s 4.\n\nIn case of SBP-2 ORB pointer write requests, 8 is arguably quicker to\nunderstand as the correct and only possible value than\nsizeof(some_datum).\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "65b2742ac002f554f6ca0c2eab84cc62400eafdf", "tree": "a13b0dc0883a5e126f2577f8061705c4c38debc0", "parents": [ "c374ab424249b6ab91b1aee7460419d3f2c321df" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 12 20:26:51 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 19 13:01:40 2010 +0200" }, "message": "firewire: \u0027add CSR_... support\u0027 addendum\n\nAdd a comment on which of the conflicting NODE_IDS specifications we\nimplement. Reduce a comment on rather irrelevant register bits that can\nall be looked up in the spec (or from now on in the code history).\nDirectly include the required indirectly included bug.h.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "c374ab424249b6ab91b1aee7460419d3f2c321df", "tree": "70da7005c955691ff436676c678cf2e9dac73aa4", "parents": [ "e91b2787d0a2e4719b016e8dec0afd2d5ab6c30f" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:41:51 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:41:51 2010 +0200" }, "message": "firewire: core: always enable cycle master packets\n\nAs part of the bus manager responsibilities, make sure that the cycle\nmaster sends cycle start packets. This is needed when the old bus\nmanager disabled the cycle master\u0027s cmstr bit and there are iso-capable\nnodes on the new bus.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "e91b2787d0a2e4719b016e8dec0afd2d5ab6c30f", "tree": "e76a3665243ed9fb7275228d9a14dcb0eb5b567a", "parents": [ "7e0e314f198d5048b74c8f0ef9f4c1c02e5ecfc9" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:40:49 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:40:49 2010 +0200" }, "message": "firewire: allocate broadcast channel in hardware\n\nOn OHCI 1.1 controllers, let the hardware allocate the broadcast channel\nautomatically. This removes a theoretical race condition directly after\na bus reset where it could be possible to read the channel allocation\nregister with channel 31 still being unallocated.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "7e0e314f198d5048b74c8f0ef9f4c1c02e5ecfc9", "tree": "513fcf8f87f6b24121d4e4eb07e1345e29d81db2", "parents": [ "4ffb7a6a066e4be4577976d1c08e237c7479770a" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:37:15 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:37:15 2010 +0200" }, "message": "firewire: core: add CSR abdicate support\n\nImplement the abdicate bit, which is required for bus manager\ncapable nodes and tested by the Base 1394 Test Suite.\n\nFinally, something to do at a command reset! :-)\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "4ffb7a6a066e4be4577976d1c08e237c7479770a", "tree": "fd4c275e8ef894d60602e1b4d1f81150e54b32f8", "parents": [ "3d1f46eb60b155c705e389ecdf313f11b4b91976" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:36:37 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:36:37 2010 +0200" }, "message": "firewire: add CSR cmstr support\n\nImplement the cmstr bit, which is required for cycle master capable\nnodes and tested for by the Base 1394 Test Suite.\n\nThis bit allows the bus master to disable cycle start packets; there are\nbus master implementations that actually do this.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "3d1f46eb60b155c705e389ecdf313f11b4b91976", "tree": "6a28e68a7372d49f172f73a2c30e1a72dd124ffa", "parents": [ "a1a1132bd83d0aea51d4f19be4b4a58a064a0131" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:35:37 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:35:37 2010 +0200" }, "message": "firewire: core: add CSR MAINT_UTILITY support\n\nImplement the MAIN_UTILITY register, which is utterly optional\nbut useful as a safe target for diagnostic read/write/broadcast\ntransactions.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "a1a1132bd83d0aea51d4f19be4b4a58a064a0131", "tree": "93352d6aa9fd8eaa0521fbeb65853cd9f485666e", "parents": [ "27a2329f8235d6ce637463f5d83e98d760ef006e" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:35:06 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:35:06 2010 +0200" }, "message": "firewire: add CSR PRIORITY_BUDGET support\n\nIf supported by the OHCI controller, implement the PRIORITY_BUDGET\nregister, which is required for nodes that can use asynchronous\npriority arbitration.\n\nTo allow the core to determine what features the lowlevel device\nsupports, add a new card driver callback.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "27a2329f8235d6ce637463f5d83e98d760ef006e", "tree": "dcc3b1a62d92d407f38744a93914742e9eed1816", "parents": [ "a48777e03ad53777ed119a5f86dd22a6c5a378ad" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:34:13 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:34:13 2010 +0200" }, "message": "firewire: add CSR BUSY_TIMEOUT support\n\nImplement the BUSY_TIMEOUT register, which is required for nodes that\nsupport retries.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "a48777e03ad53777ed119a5f86dd22a6c5a378ad", "tree": "651a3fb567a2b9d6ab111ff07415d87397ba8202", "parents": [ "9ab5071cd4a16001e4ba790172a7da5e4172462b" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:33:07 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:33:07 2010 +0200" }, "message": "firewire: add CSR BUS_TIME support\n\nImplement the BUS_TIME register, which is required for cycle master\ncapable nodes and tested for by the Base 1393 Test Suite. Even when\nthere is not yet bus master initialization support, this register allows\nus to work together with other bus masters.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "9ab5071cd4a16001e4ba790172a7da5e4172462b", "tree": "bb1ba598d4f2df93b661411c833b60bb953e3390", "parents": [ "8e4b50f94e8c1435a3e0ece42b7f97bc857d0145" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:26:48 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:26:48 2010 +0200" }, "message": "firewire: add CSR CYCLE_TIME write support\n\nThe specification requires that CYCLE_TIME is writable so that it can be\ninitialized, so we better implement it.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "8e4b50f94e8c1435a3e0ece42b7f97bc857d0145", "tree": "aad5b85a0bf67b5a7bbd80e1dad2d2a2d9b85470", "parents": [ "446eba0d6896787b2f02f7a665838d32aa7b9d3f" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:26:28 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:26:28 2010 +0200" }, "message": "firewire: core: add CSR SPLIT_TIMEOUT support\n\nImplement the SPLIT_TIMEOUT registers. Besides being required by the\nspec, this is desirable for some IIDC devices and necessary for many\naudio devices to be able to increase the timeout from userspace.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "446eba0d6896787b2f02f7a665838d32aa7b9d3f", "tree": "a4428929ea4f7512cf9fd77f60195ce6bc8af08e", "parents": [ "506f1a31932747f56a5029d5b3c14b1b68f41ccc" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:25:46 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:25:46 2010 +0200" }, "message": "firewire: core: add CSR RESET_START support\n\nThis implements the RESET_START register (as a dummy) to make the Base\n1394 Test Suite happy.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "506f1a31932747f56a5029d5b3c14b1b68f41ccc", "tree": "04dfb94980fc95cfc2c458fc62e6b77996c8145b", "parents": [ "60d32970c5a32e8c4f340a9e41993759ad658ef2" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:25:19 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:25:19 2010 +0200" }, "message": "firewire: add CSR NODE_IDS support\n\nThe NODE_IDS register, and especially its bus_id field, is quite\nuseless because 1394.1 requires that the bus_id field always stays\n0x3ff. However, the 1394 specification requires this register on all\ntransaction capable nodes, and the Base 1394 Test Suite tests for it,\nso we better implement it.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "60d32970c5a32e8c4f340a9e41993759ad658ef2", "tree": "a7f60c934ae250f0291622cb02540638ef7cbf6d", "parents": [ "3e07ec0eee1662f89e57f84aff625065beb2b209" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:24:35 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:24:35 2010 +0200" }, "message": "firewire: add read_csr_reg driver callback\n\nTo prepare for the following additions of more OHCI-implemented CSR\nregisters, replace the get_cycle_time driver callback with a generic\nCSR register callback.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "3e07ec0eee1662f89e57f84aff625065beb2b209", "tree": "ef164bc604c9b22d21983ae90dbe24636264db9d", "parents": [ "bda3b8a1faf209a98063ccd77d6833a2bb0fc77e" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:24:03 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:24:03 2010 +0200" }, "message": "firewire: core: add CSR STATE_CLEAR/STATE_SET support\n\nThe state registers are zero and read-only in this implementation, so\nthey are not of much use. However, the specification requires that they\nare present for transaction capable nodes, and the Base 1394 Test Suite\ntests for them, so we better implement them.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "bda3b8a1faf209a98063ccd77d6833a2bb0fc77e", "tree": "1f7146d8951a5addac9932a071a6f0e4a4d97182", "parents": [ "153e3979201b76dbd5788f032fb683e95121e159" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:23:28 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:23:28 2010 +0200" }, "message": "firewire: core: retry on local errors in bus manager election\n\nWhen the candidate bus manager fails to do the lock request with which\nit tries to become bus manager, it assumes that the current IRM is not\nactually IRM capable and forces itself to become root. However, if that\nlock request failed because the local node itself was not able to send\nit, then we cannot blame the current IRM and should not steal its\nrootness.\n\nIn this case, RCODE_SEND_ERROR is likely to indicate a temporary error\ncondition such as exhausted tlabels or low memory, so we better try\nagain later.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "153e3979201b76dbd5788f032fb683e95121e159", "tree": "aae0a3a5dcbc0eb815e7030176b03c84da8792b9", "parents": [ "f9c70f9129f2d88645c3a26711302a7f6ba9afd0" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:22:07 2010 +0200" }, "committer": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Thu Jun 10 08:22:07 2010 +0200" }, "message": "firewire: ohci: speed up PHY register accesses\n\nMost PHY chips, when idle, can complete a register access in the time\nneeded for two or three PCI read transactions; bigger delays occur only\nwhen data is currently being moved over the link/PHY interface. So if\nwe busy-wait a few times when waiting for the register access to finish,\nit is likely that we can finish without having to sleep.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n" }, { "commit": "f9c70f9129f2d88645c3a26711302a7f6ba9afd0", "tree": "5a50106146109d2a1744a809a3fef7f9d8bee7d1", "parents": [ "a10c0ce76098857b899505d05de9f2e13ddf7a7a" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 05 20:32:50 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jun 09 19:42:18 2010 +0200" }, "message": "firewire: core: trivial fix for warning strings\n\nWARN\u0027s format string argument should not carry a printk level prefix.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "a10c0ce76098857b899505d05de9f2e13ddf7a7a", "tree": "130592c6baaff2e38dd813448337dded1ee1645b", "parents": [ "262444eecce40950af19ea4d75a3dc03b3c07283" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Wed May 19 08:28:32 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jun 09 19:42:18 2010 +0200" }, "message": "firewire: check cdev response length\n\nAdd a check that the data length in the SEND_RESPONSE ioctl is correct.\nIncidentally, this also fixes the previously wrong response length of\nsoftware-handled lock requests.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "262444eecce40950af19ea4d75a3dc03b3c07283", "tree": "87b465d2bdc2cc600807910d372e68ba668bda8a", "parents": [ "148c7866c31d93f8c79366189075f5a26ad4556c" ], "author": { "name": "Clemens Ladisch", "email": "clemens@ladisch.de", "time": "Sat Jun 05 12:31:25 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jun 09 19:42:18 2010 +0200" }, "message": "firewire: ohci: add MSI support\n\nThis patch adds support for message-signaled interrupts.\n\nAny native PCI-Express OHCI controller should support MSI, but most are\njust PCI cores behind a PCI-E/PCI bridge. The only chips that are known\nto claim to support MSI are the Lucent/Agere/LSI FW643 and the VIA\nVT6315, none of which I have been able to test.\n\nDue to the high level of trust I have in the competence of these and any\nfuture chip makers, I thought it a good idea to add a disable-MSI quirk.\n\nSigned-off-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\n\nTested Agere FW643 rev 07 [11c1:5901] and JMicron JMB381 [197b:2380].\nAdded a quirks list entry for JMB38X since it kept its count of MSI\nevents consistently at zero.\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "148c7866c31d93f8c79366189075f5a26ad4556c", "tree": "2b82bfd64fcdf04a2fbb6bfca78e27c9c609110f", "parents": [ "d8c1fa4af0f311363d9f9cf1014b11d31a99ff10" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sat Jun 05 11:46:49 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jun 09 19:42:18 2010 +0200" }, "message": "firewire: ohci: do not enable interrupts without the handler\n\nOn 26 Apr 2010, Clemens Ladisch wrote:\n\u003e In theory, none of the interrupts should occur before the link is\n\u003e enabled. In practice, I\u0027d rather make sure to not set the master\n\u003e interrupt enable bit until we have installed the interrupt handler.\n\nand proposed to move OHCI1394_masterIntEnable out of the present\nreg_write() into a new one before the HCControl.linkEnable reg_write().\n\nWhy not defer setting /all/ of the bits until right before linkEnable?\n\nReviewed-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" }, { "commit": "10389536742cefbedecb67a5b2906f155cf3a1c3", "tree": "cf0dbc030578e267e7f9fe1349083c3af3090f19", "parents": [ "67a3e12b05e055c0415c556a315a3d3eb637e29e" ], "author": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Sun May 30 19:43:52 2010 +0200" }, "committer": { "name": "Stefan Richter", "email": "stefanr@s5r6.in-berlin.de", "time": "Wed Jun 02 19:48:13 2010 +0200" }, "message": "firewire: core: check for 1394a compliant IRM, fix inaccessibility of Sony camcorder\n\nPer IEEE 1394 clause 8.4.2.3, a contender for the IRM role shall check\nwhether the current IRM complies to 1394a-2000 or later. If not force a\ncompliant node (e.g. itself) to become IRM. This was implemented in the\nolder ieee1394 driver but not yet in firewire-core.\n\nAn older Sony camcorder (Sony DCR-TRV25) which implements 1394-1995 IRM\nbut neither 1394a-2000 IRM nor BM was now found to cause an\ninteroperability bug:\n - Camcorder becomes root node when plugged in, hence gets IRM role.\n - firewire-core successfully contends for BM role, proceeds to perform\n gap count optimization and resets the bus.\n - Sony camcorder ignores presence of a BM (against the spec, this is\n a firmware bug), performs its idea of gap count optimization and\n resets the bus.\n - Preceding two steps are repeated endlessly, bus never settles,\n regular I/O is practically impossible.\nhttp://thread.gmane.org/gmane.linux.kernel.firewire.user/3913\n\nThis is an interoperability regression from the old to the new drivers.\nFix it indirectly by adding the 1394a IRM check. The spec suggests\nthree and a half methods to determine 1394a compliance of a remote IRM;\nwe choose the method of testing the Config_ROM.Bus_Info.generation\nfield. This is data that firewire-core should have readily available at\nthis point, i.e. does not require extra I/O.\n\nReported-by: Clemens Ladisch \u003cclemens@ladisch.de\u003e (missing 1394a check)\nReported-by: H. S. \u003chs.samix@gmail.com\u003e (issue with Sony DCR-TRV25)\nTested-by: H. S. \u003chs.samix@gmail.com\u003e\n\nCc: \u003cstable@kernel.org\u003e # .32.x and newer\n\nSigned-off-by: Stefan Richter \u003cstefanr@s5r6.in-berlin.de\u003e\n" } ], "next": "55ddf14b04bfe5afaab892a8fb12164b803f1dd5" }