ChangeLog.megaraid 23 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614
  1. Release Date : Thu Nov 16 15:32:35 EST 2006 -
  2. Sumant Patro <sumant.patro@lsi.com>
  3. Current Version : 2.20.5.1 (scsi module), 2.20.2.6 (cmm module)
  4. Older Version : 2.20.4.9 (scsi module), 2.20.2.6 (cmm module)
  5. 1. Changes in Initialization to fix kdump failure.
  6. Send SYNC command on loading.
  7. This command clears the pending commands in the adapter
  8. and re-initialize its internal RAID structure.
  9. Without this change, megaraid driver either panics or fails to
  10. initialize the adapter during kdump's second kernel boot
  11. if there are pending commands or interrupts from other devices
  12. sharing the same IRQ.
  13. 2. Authors email-id domain name changed from lsil.com to lsi.com.
  14. Also modified the MODULE_AUTHOR to megaraidlinux@lsi.com
  15. Release Date : Fri May 19 09:31:45 EST 2006 - Seokmann Ju <sju@lsil.com>
  16. Current Version : 2.20.4.9 (scsi module), 2.20.2.6 (cmm module)
  17. Older Version : 2.20.4.8 (scsi module), 2.20.2.6 (cmm module)
  18. 1. Fixed a bug in megaraid_init_mbox().
  19. Customer reported "garbage in file on x86_64 platform".
  20. Root Cause: the driver registered controllers as 64-bit DMA capable
  21. for those which are not support it.
  22. Fix: Made change in the function inserting identification machanism
  23. identifying 64-bit DMA capable controllers.
  24. > -----Original Message-----
  25. > From: Vasily Averin [mailto:vvs@sw.ru]
  26. > Sent: Thursday, May 04, 2006 2:49 PM
  27. > To: linux-scsi@vger.kernel.org; Kolli, Neela; Mukker, Atul;
  28. > Ju, Seokmann; Bagalkote, Sreenivas;
  29. > James.Bottomley@SteelEye.com; devel@openvz.org
  30. > Subject: megaraid_mbox: garbage in file
  31. >
  32. > Hello all,
  33. >
  34. > I've investigated customers claim on the unstable work of
  35. > their node and found a
  36. > strange effect: reading from some files leads to the
  37. > "attempt to access beyond end of device" messages.
  38. >
  39. > I've checked filesystem, memory on the node, motherboard BIOS
  40. > version, but it
  41. > does not help and issue still has been reproduced by simple
  42. > file reading.
  43. >
  44. > Reproducer is simple:
  45. >
  46. > echo 0xffffffff >/proc/sys/dev/scsi/logging_level ;
  47. > cat /vz/private/101/root/etc/ld.so.cache >/tmp/ttt ;
  48. > echo 0 >/proc/sys/dev/scsi/logging
  49. >
  50. > It leads to the following messages in dmesg
  51. >
  52. > sd_init_command: disk=sda, block=871769260, count=26
  53. > sda : block=871769260
  54. > sda : reading 26/26 512 byte blocks.
  55. > scsi_add_timer: scmd: f79ed980, time: 7500, (c02b1420)
  56. > sd 0:1:0:0: send 0xf79ed980 sd 0:1:0:0:
  57. > command: Read (10): 28 00 33 f6 24 ac 00 00 1a 00
  58. > buffer = 0xf7cfb540, bufflen = 13312, done = 0xc0366b40,
  59. > queuecommand 0xc0344010
  60. > leaving scsi_dispatch_cmnd()
  61. > scsi_delete_timer: scmd: f79ed980, rtn: 1
  62. > sd 0:1:0:0: done 0xf79ed980 SUCCESS 0 sd 0:1:0:0:
  63. > command: Read (10): 28 00 33 f6 24 ac 00 00 1a 00
  64. > scsi host busy 1 failed 0
  65. > sd 0:1:0:0: Notifying upper driver of completion (result 0)
  66. > sd_rw_intr: sda: res=0x0
  67. > 26 sectors total, 13312 bytes done.
  68. > use_sg is 4
  69. > attempt to access beyond end of device
  70. > sda6: rw=0, want=1044134458, limit=951401367
  71. > Buffer I/O error on device sda6, logical block 522067228
  72. > attempt to access beyond end of device
  73. 2. When INQUIRY with EVPD bit set issued to the MegaRAID controller,
  74. system memory gets corrupted.
  75. Root Cause: MegaRAID F/W handle the INQUIRY with EVPD bit set
  76. incorrectly.
  77. Fix: MegaRAID F/W has fixed the problem and being process of release,
  78. soon. Meanwhile, driver will filter out the request.
  79. 3. One of member in the data structure of the driver leads unaligne
  80. issue on 64-bit platform.
  81. Customer reporeted "kernel unaligned access addrss" issue when
  82. application communicates with MegaRAID HBA driver.
  83. Root Cause: in uioc_t structure, one of member had misaligned and it
  84. led system to display the error message.
  85. Fix: A patch submitted to community from following folk.
  86. > -----Original Message-----
  87. > From: linux-scsi-owner@vger.kernel.org
  88. > [mailto:linux-scsi-owner@vger.kernel.org] On Behalf Of Sakurai Hiroomi
  89. > Sent: Wednesday, July 12, 2006 4:20 AM
  90. > To: linux-scsi@vger.kernel.org; linux-kernel@vger.kernel.org
  91. > Subject: Re: Help: strange messages from kernel on IA64 platform
  92. >
  93. > Hi,
  94. >
  95. > I saw same message.
  96. >
  97. > When GAM(Global Array Manager) is started, The following
  98. > message output.
  99. > kernel: kernel unaligned access to 0xe0000001fe1080d4,
  100. > ip=0xa000000200053371
  101. >
  102. > The uioc structure used by ioctl is defined by packed,
  103. > the allignment of each member are disturbed.
  104. > In a 64 bit structure, the allignment of member doesn't fit 64 bit
  105. > boundary. this causes this messages.
  106. > In a 32 bit structure, we don't see the message because the allinment
  107. > of member fit 32 bit boundary even if packed is specified.
  108. >
  109. > patch
  110. > I Add 32 bit dummy member to fit 64 bit boundary. I tested.
  111. > We confirmed this patch fix the problem by IA64 server.
  112. >
  113. > **************************************************************
  114. > ****************
  115. > --- linux-2.6.9/drivers/scsi/megaraid/megaraid_ioctl.h.orig
  116. > 2006-04-03 17:13:03.000000000 +0900
  117. > +++ linux-2.6.9/drivers/scsi/megaraid/megaraid_ioctl.h
  118. > 2006-04-03 17:14:09.000000000 +0900
  119. > @@ -132,6 +132,10 @@
  120. > /* Driver Data: */
  121. > void __user * user_data;
  122. > uint32_t user_data_len;
  123. > +
  124. > + /* 64bit alignment */
  125. > + uint32_t pad_0xBC;
  126. > +
  127. > mraid_passthru_t __user *user_pthru;
  128. >
  129. > mraid_passthru_t *pthru32;
  130. > **************************************************************
  131. > ****************
  132. Release Date : Mon Apr 11 12:27:22 EST 2006 - Seokmann Ju <sju@lsil.com>
  133. Current Version : 2.20.4.8 (scsi module), 2.20.2.6 (cmm module)
  134. Older Version : 2.20.4.7 (scsi module), 2.20.2.6 (cmm module)
  135. 1. Fixed a bug in megaraid_reset_handler().
  136. Customer reported "Unable to handle kernel NULL pointer dereference
  137. at virtual address 00000000" when system goes to reset condition
  138. for some reason. It happened randomly.
  139. Root Cause: in the megaraid_reset_handler(), there is possibility not
  140. returning pending packets in the pend_list if there are multiple
  141. pending packets.
  142. Fix: Made the change in the driver so that it will return all packets
  143. in the pend_list.
  144. 2. Added change request.
  145. As found in the following URL, rmb() only didn't help the
  146. problem. I had to increase the loop counter to 0xFFFFFF. (6 F's)
  147. http://marc.theaimsgroup.com/?l=linux-scsi&m=110971060502497&w=2
  148. I attached a patch for your reference, too.
  149. Could you check and get this fix in your driver?
  150. Best Regards,
  151. Jun'ichi Nomura
  152. Release Date : Fri Nov 11 12:27:22 EST 2005 - Seokmann Ju <sju@lsil.com>
  153. Current Version : 2.20.4.7 (scsi module), 2.20.2.6 (cmm module)
  154. Older Version : 2.20.4.6 (scsi module), 2.20.2.6 (cmm module)
  155. 1. Sorted out PCI IDs to remove megaraid support overlaps.
  156. Based on the patch from Daniel, sorted out PCI IDs along with
  157. character node name change from 'megadev' to 'megadev_legacy' to avoid
  158. conflict.
  159. ---
  160. Hopefully we'll be getting the build restriction zapped much sooner,
  161. but we should also be thinking about totally removing the hardware
  162. support overlap in the megaraid drivers.
  163. This patch pencils in a date of Feb 06 for this, and performs some
  164. printk abuse in hope that existing legacy users might pick up on what's
  165. going on.
  166. Signed-off-by: Daniel Drake <dsd@gentoo.org>
  167. ---
  168. 2. Fixed a issue: megaraid always fails to reset handler.
  169. ---
  170. I found that the megaraid driver always fails to reset the
  171. adapter with the following message:
  172. megaraid: resetting the host...
  173. megaraid mbox: reset sequence completed successfully
  174. megaraid: fast sync command timed out
  175. megaraid: reservation reset failed
  176. when the "Cluster mode" of the adapter BIOS is enabled.
  177. So, whenever the reset occurs, the adapter goes to
  178. offline and just become unavailable.
  179. Jun'ichi Nomura [mailto:jnomura@mtc.biglobe.ne.jp]
  180. ---
  181. Release Date : Mon Mar 07 12:27:22 EST 2005 - Seokmann Ju <sju@lsil.com>
  182. Current Version : 2.20.4.6 (scsi module), 2.20.2.6 (cmm module)
  183. Older Version : 2.20.4.5 (scsi module), 2.20.2.5 (cmm module)
  184. 1. Added IOCTL backward compatibility.
  185. Convert megaraid_mm driver to new compat_ioctl entry points.
  186. I don't have easy access to hardware, so only compile tested.
  187. - Signed-off-by:Andi Kleen <ak@muc.de>
  188. 2. megaraid_mbox fix: wrong order of arguments in memset()
  189. That, BTW, shows why cross-builds are useful-the only indication of
  190. problem had been a new warning showing up in sparse output on alpha
  191. build (number of exceeding 256 got truncated).
  192. - Signed-off-by: Al Viro
  193. <viro@parcelfarce.linux.theplanet.co.uk>
  194. 3. Convert pci_module_init to pci_register_driver
  195. Convert from pci_module_init to pci_register_driver
  196. (from:http://kernelnewbies.org/KernelJanitors/TODO)
  197. - Signed-off-by: Domen Puncer <domen@coderock.org>
  198. 4. Use the pre defined DMA mask constants from dma-mapping.h
  199. Use the DMA_{64,32}BIT_MASK constants from dma-mapping.h when calling
  200. pci_set_dma_mask() or pci_set_consistend_dma_mask(). See
  201. http://marc.theaimsgroup.com/?t=108001993000001&r=1&w=2 for more
  202. details.
  203. Signed-off-by: Tobias Klauser <tklauser@nuerscht.ch>
  204. Signed-off-by: Domen Puncer <domen@coderock.org>
  205. 5. Remove SSID checking for Dobson, Lindsay, and Verde based products.
  206. Checking the SSVID/SSID for controllers which have Dobson, Lindsay,
  207. and Verde is unnecessary because device ID has been assigned by LSI
  208. and it is unique value. So, all controllers with these IOPs have to be
  209. supported by the driver regardless SSVID/SSID.
  210. 6. Date Thu, 27 Jan 2005 04:31:09 +0100
  211. From Herbert Poetzl <>
  212. Subject RFC: assert_spin_locked() for 2.6
  213. Greetings!
  214. overcautious programming will kill your kernel ;)
  215. ever thought about checking a spin_lock or even
  216. asserting that it must be held (maybe just for
  217. spinlock debugging?) ...
  218. there are several checks present in the kernel
  219. where somebody does a variation on the following:
  220. BUG_ON(!spin_is_locked(&some_lock));
  221. so what's wrong about that? nothing, unless you
  222. compile the code with CONFIG_DEBUG_SPINLOCK but
  223. without CONFIG_SMP ... in which case the BUG()
  224. will kill your kernel ...
  225. maybe it's not advised to make such assertions,
  226. but here is a solution which works for me ...
  227. (compile tested for sh, x86_64 and x86, boot/run
  228. tested for x86 only)
  229. best,
  230. Herbert
  231. - Herbert Poetzl <herbert@13thfloor.at>, Thu, 27 Jan 2005
  232. Release Date : Thu Feb 03 12:27:22 EST 2005 - Seokmann Ju <sju@lsil.com>
  233. Current Version : 2.20.4.5 (scsi module), 2.20.2.5 (cmm module)
  234. Older Version : 2.20.4.4 (scsi module), 2.20.2.4 (cmm module)
  235. 1. Modified name of two attributes in scsi_host_template.
  236. On Wed, 2005-02-02 at 10:56 -0500, Ju, Seokmann wrote:
  237. > + .sdev_attrs = megaraid_device_attrs,
  238. > + .shost_attrs = megaraid_class_device_attrs,
  239. These are, perhaps, slightly confusing names.
  240. The terms device and class_device have well defined meanings in the
  241. generic device model, neither of which is what you mean here.
  242. Why not simply megaraid_sdev_attrs and megaraid_shost_attrs?
  243. Other than this, it looks fine to me too.
  244. Release Date : Thu Jan 27 00:01:03 EST 2005 - Atul Mukker <atulm@lsil.com>
  245. Current Version : 2.20.4.4 (scsi module), 2.20.2.5 (cmm module)
  246. Older Version : 2.20.4.3 (scsi module), 2.20.2.4 (cmm module)
  247. 1. Bump up the version of scsi module due to its conflict.
  248. Release Date : Thu Jan 21 00:01:03 EST 2005 - Atul Mukker <atulm@lsil.com>
  249. Current Version : 2.20.4.3 (scsi module), 2.20.2.5 (cmm module)
  250. Older Version : 2.20.4.2 (scsi module), 2.20.2.4 (cmm module)
  251. 1. Remove driver ioctl for logical drive to scsi address translation and
  252. replace with the sysfs attribute. To remove drives and change
  253. capacity, application shall now use the device attribute to get the
  254. logical drive number for a scsi device. For adding newly created
  255. logical drives, class device attribute would be required to uniquely
  256. identify each controller.
  257. - Atul Mukker <atulm@lsil.com>
  258. "James, I've been thinking about this a little more, and you may be on
  259. to something here. Let each driver add files as such:"
  260. - Matt Domsch <Matt_Domsch@dell.com>, 12.15.2004
  261. linux-scsi mailing list
  262. "Then, if you simply publish your LD number as an extra parameter of
  263. the device, you can look through /sys to find it."
  264. - James Bottomley <James.Bottomley@SteelEye.com>, 01.03.2005
  265. linux-scsi mailing list
  266. "I don't see why not ... it's your driver, you can publish whatever
  267. extra information you need as scsi_device attributes; that was one of
  268. the designs of the extensible attribute system."
  269. - James Bottomley <James.Bottomley@SteelEye.com>, 01.06.2005
  270. linux-scsi mailing list
  271. 2. Add AMI megaraid support - Brian King <brking@charter.net>
  272. PCI_VENDOR_ID_AMI, PCI_DEVICE_ID_AMI_MEGARAID3,
  273. PCI_VENDOR_ID_AMI, PCI_SUBSYS_ID_PERC3_DC,
  274. 3. Make some code static - Adrian Bunk <bunk@stusta.de>
  275. Date: Mon, 15 Nov 2004 03:14:57 +0100
  276. The patch below makes some needlessly global code static.
  277. -wait_queue_head_t wait_q;
  278. +static wait_queue_head_t wait_q;
  279. Signed-off-by: Adrian Bunk <bunk@stusta.de>
  280. 4. Added NEC ROMB support - NEC MegaRAID PCI Express ROMB controller
  281. PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_MEGARAID_NEC_ROMB_2E,
  282. PCI_SUBSYS_ID_NEC, PCI_SUBSYS_ID_MEGARAID_NEC_ROMB_2E,
  283. 5. Fixed Tape drive issue : For any Direct CDB command to physical device
  284. including tape, timeout value set by driver was 10 minutes. With this
  285. value, most of command will return within timeout. However, for those
  286. command like ERASE or FORMAT, it takes more than an hour depends on
  287. capacity of the device and the command could be terminated before it
  288. completes.
  289. To address this issue, the 'timeout' field in the DCDB command will
  290. have NO TIMEOUT (i.e., 4) value as its timeout on DCDB command.
  291. Release Date : Thu Dec 9 19:10:23 EST 2004
  292. - Sreenivas Bagalkote <sreenib@lsil.com>
  293. Current Version : 2.20.4.2 (scsi module), 2.20.2.4 (cmm module)
  294. Older Version : 2.20.4.1 (scsi module), 2.20.2.3 (cmm module)
  295. i. Introduced driver ioctl that returns scsi address for a given ld.
  296. "Why can't the existing sysfs interfaces be used to do this?"
  297. - Brian King (brking@us.ibm.com)
  298. "I've looked into solving this another way, but I cannot see how
  299. to get this driver-private mapping of logical drive number-> HCTL
  300. without putting code something like this into the driver."
  301. "...and by providing a mapping a function to userspace, the driver
  302. is free to change its mapping algorithm in the future if necessary .."
  303. - Matt Domsch (Matt_Domsch@dell.com)
  304. Release Date : Thu Dec 9 19:02:14 EST 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
  305. Current Version : 2.20.4.1 (scsi module), 2.20.2.3 (cmm module)
  306. Older Version : 2.20.4.1 (scsi module), 2.20.2.2 (cmm module)
  307. i. Fix a bug in kioc's dma buffer deallocation
  308. Release Date : Thu Nov 4 18:24:56 EST 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
  309. Current Version : 2.20.4.1 (scsi module), 2.20.2.2 (cmm module)
  310. Older Version : 2.20.4.0 (scsi module), 2.20.2.1 (cmm module)
  311. i. Handle IOCTL cmd timeouts more properly.
  312. ii. pci_dma_sync_{sg,single}_for_cpu was introduced into megaraid_mbox
  313. incorrectly (instead of _for_device). Changed to appropriate
  314. pci_dma_sync_{sg,single}_for_device.
  315. Release Date : Wed Oct 06 11:15:29 EDT 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
  316. Current Version : 2.20.4.0 (scsi module), 2.20.2.1 (cmm module)
  317. Older Version : 2.20.4.0 (scsi module), 2.20.2.0 (cmm module)
  318. i. Remove CONFIG_COMPAT around register_ioctl32_conversion
  319. Release Date : Mon Sep 27 22:15:07 EDT 2004 - Atul Mukker <atulm@lsil.com>
  320. Current Version : 2.20.4.0 (scsi module), 2.20.2.0 (cmm module)
  321. Older Version : 2.20.3.1 (scsi module), 2.20.2.0 (cmm module)
  322. i. Fix data corruption. Because of a typo in the driver, the IO packets
  323. were wrongly shared by the ioctl path. This causes a whole IO command
  324. to be replaced by an incoming ioctl command.
  325. Release Date : Tue Aug 24 09:43:35 EDT 2004 - Atul Mukker <atulm@lsil.com>
  326. Current Version : 2.20.3.1 (scsi module), 2.20.2.0 (cmm module)
  327. Older Version : 2.20.3.0 (scsi module), 2.20.2.0 (cmm module)
  328. i. Function reordering so that inline functions are defined before they
  329. are actually used. It is now mandatory for GCC 3.4.1 (current stable)
  330. Declare some heavy-weight functions to be non-inlined,
  331. megaraid_mbox_build_cmd, megaraid_mbox_runpendq,
  332. megaraid_mbox_prepare_pthru, megaraid_mbox_prepare_epthru,
  333. megaraid_busywait_mbox
  334. - Andrew Morton, 08.19.2004
  335. linux-scsi mailing list
  336. "Something else to clean up after inclusion: every instance of an
  337. inline function is actually rendered as a full function call, because
  338. the function is always used before it is defined. Atul, please
  339. re-arrange the code to eliminate the need for most (all) of the
  340. function prototypes at the top of each file, and define (not just
  341. declare with a prototype) each inline function before its first use"
  342. - Matt Domsch <Matt_Domsch@dell.com>, 07.27.2004
  343. linux-scsi mailing list
  344. ii. Display elapsed time (countdown) while waiting for FW to boot.
  345. iii. Module compilation reorder in Makefile so that unresolved symbols do
  346. not occur when driver is compiled non-modular.
  347. Patrick J. LoPresti <patl@users.sourceforge.net>, 8.22.2004
  348. linux-scsi mailing list
  349. Release Date : Thu Aug 19 09:58:33 EDT 2004 - Atul Mukker <atulm@lsil.com>
  350. Current Version : 2.20.3.0 (scsi module), 2.20.2.0 (cmm module)
  351. Older Version : 2.20.2.0 (scsi module), 2.20.1.0 (cmm module)
  352. i. When copying the mailbox packets, copy only first 14 bytes (for 32-bit
  353. mailboxes) and only first 22 bytes (for 64-bit mailboxes). This is to
  354. avoid getting the stale values for busy bit. We want to set the busy
  355. bit just before issuing command to the FW.
  356. ii. In the reset handling, if the reseted command is not owned by the
  357. driver, do not (wrongly) print information for the "attached" driver
  358. packet.
  359. iii. Have extended wait when issuing command in synchronous mode. This is
  360. required for the cases where the option ROM is disabled and there is
  361. no BIOS to start the controller. The FW starts to boot after receiving
  362. the first command from the driver. The current driver has 1 second
  363. timeout for the synchronous commands, which is far less than what is
  364. actually required. We now wait up to MBOX_RESET_TIME (180 seconds) for
  365. FW boot process.
  366. iv. In megaraid_mbox_product_info, clear the mailbox contents completely
  367. before preparing the command for inquiry3. This is to ensure that the
  368. FW does not get junk values in the command.
  369. v. Do away with the redundant LSI_CONFIG_COMPAT redefinition for
  370. CONFIG_COMPAT. Replace <asm/ioctl32.h> with <linux/ioctl32.h>
  371. - James Bottomley <James.Bottomley@SteelEye.com>, 08.17.2004
  372. linux-scsi mailing list
  373. vi. Add support for 64-bit applications. Current drivers assume only
  374. 32-bit applications, even on 64-bit platforms. Use the "data" and
  375. "buffer" fields of the mimd_t structure, instead of embedded 32-bit
  376. addresses in application mailbox and passthru structures.
  377. vii. Move the function declarations for the management module from
  378. megaraid_mm.h to megaraid_mm.c
  379. - Andrew Morton, 08.19.2004
  380. linux-scsi mailing list
  381. viii. Change default values for MEGARAID_NEWGEN, MEGARAID_MM, and
  382. MEGARAID_MAILBOX to 'n' in Kconfig.megaraid
  383. - Andrew Morton, 08.19.2004
  384. linux-scsi mailing list
  385. ix. replace udelay with msleep
  386. x. Typos corrected in comments and whitespace adjustments, explicit
  387. grouping of expressions.
  388. Release Date : Fri Jul 23 15:22:07 EDT 2004 - Atul Mukker <atulm@lsil.com>
  389. Current Version : 2.20.2.0 (scsi module), 2.20.1.0 (cmm module)
  390. Older Version : 2.20.1.0 (scsi module), 2.20.0.0 (cmm module)
  391. i. Add PCI ids for Acer ROMB 2E solution
  392. ii. Add PCI ids for I4
  393. iii. Typo corrected for subsys id for megaraid sata 300-4x
  394. iv. Remove yield() while mailbox handshake in synchronous commands
  395. "My other main gripe is things like this:
  396. + // wait for maximum 1 second for status to post
  397. + for (i = 0; i < 40000; i++) {
  398. + if (mbox->numstatus != 0xFF) break;
  399. + udelay(25); yield();
  400. + }
  401. which litter the driver. Use of yield() in drivers is deprecated."
  402. - James Bottomley <James.Bottomley@SteelEye.com>, 07.14.2004
  403. linux-scsi mailing list
  404. v. Remove redundant __megaraid_busywait_mbox routine
  405. vi. Fix bug in the management module, which causes a system lockup when the
  406. IO module is loaded and then unloaded, followed by executing any
  407. management utility. The current version of management module does not
  408. handle the adapter unregister properly.
  409. Specifically, it still keeps a reference to the unregistered
  410. controllers. To avoid this, the static array adapters has been
  411. replaced by a dynamic list, which gets updated every time an adapter
  412. is added or removed.
  413. Also, during unregistration of the IO module, the resources are
  414. now released in the exact reverse order of the allocation time
  415. sequence.
  416. Release Date : Fri Jun 25 18:58:43 EDT 2004 - Atul Mukker <atulm@lsil.com>
  417. Current Version : 2.20.1.0
  418. Older Version : megaraid 2.20.0.1
  419. i. Stale list pointer in adapter causes kernel panic when module
  420. megaraid_mbox is unloaded
  421. Release Date : Thu Jun 24 20:37:11 EDT 2004 - Atul Mukker <atulm@lsil.com>
  422. Current Version : 2.20.0.1
  423. Older Version : megaraid 2.20.0.00
  424. i. Modules are not 'y' by default, but depend on current definition of
  425. SCSI & PCI.
  426. ii. Redundant structure mraid_driver_t removed.
  427. iii. Miscellaneous indentation and goto/label fixes.
  428. - Christoph Hellwig <hch@infradead.org>, 06.24.2004 linux-scsi
  429. iv. scsi_host_put(), do just before completing HBA shutdown.
  430. Release Date : Mon Jun 21 19:53:54 EDT 2004 - Atul Mukker <atulm@lsil.com>
  431. Current Version : 2.20.0.0
  432. Older Version : megaraid 2.20.0.rc2 and 2.00.3
  433. i. Independent module to interact with userland applications and
  434. multiplex command to low level RAID module(s).
  435. "Shared code in a third module, a "library module", is an acceptable
  436. solution. modprobe automatically loads dependent modules, so users
  437. running "modprobe driver1" or "modprobe driver2" would automatically
  438. load the shared library module."
  439. - Jeff Garzik <jgarzik@pobox.com> 02.25.2004 LKML
  440. "As Jeff hinted, if your userspace<->driver API is consistent between
  441. your new MPT-based RAID controllers and your existing megaraid driver,
  442. then perhaps you need a single small helper module (lsiioctl or some
  443. better name), loaded by both mptraid and megaraid automatically, which
  444. handles registering the /dev/megaraid node dynamically. In this case,
  445. both mptraid and megaraid would register with lsiioctl for each
  446. adapter discovered, and lsiioctl would essentially be a switch,
  447. redirecting userspace tool ioctls to the appropriate driver."
  448. - Matt Domsch <Matt_Domsch@dell.com> 02.25.2004 LKML
  449. ii. Remove C99 initializations from pci_device id.
  450. "pci_id_table_g would be much more readable when not using C99
  451. initializers.
  452. PCI table doesn't change, there's lots of users that prefer the more
  453. readable variant. And it's really far less and much easier to grok
  454. lines without C99 initializers."
  455. - Christoph Hellwig <hch@infradead.org>, 05.28.2004 linux-scsi
  456. iii. Many fixes as suggested by Christoph Hellwig <hch@infradead.org> on
  457. linux-scsi, 05.28.2004
  458. iv. We now support up to 32 parallel ioctl commands instead of current 1.
  459. There is a conscious effort to let memory allocation not fail for ioctl
  460. commands.
  461. v. Do away with internal memory management. Use pci_pool_(create|alloc)
  462. instead.
  463. vi. Kill tasklet when unloading the driver.
  464. vii. Do not use "host_lock', driver has fine-grain locks now to protect all
  465. data structures.
  466. viii. Optimize the build scatter-gather list routine. The callers already
  467. know the data transfer address and length.
  468. ix. Better implementation of error handling and recovery. Driver now
  469. performs extended errors recovery for instances like scsi cable pull.
  470. x. Disassociate the management commands with an overlaid scsi command.
  471. Driver now treats the management packets as special packets and has a
  472. dedicated callback routine.