Documentation/scsi/ChangeLog.megaraid

   1 Release Date    : Fri May 19 09:31:45 EST 2006 - Seokmann Ju <sju@lsil.com>
   2 Current Version : 2.20.4.9 (scsi module), 2.20.2.6 (cmm module)
   3 Older Version   : 2.20.4.8 (scsi module), 2.20.2.6 (cmm module)
   4
   5 1.      Fixed a bug in megaraid_init_mbox().
   6         Customer reported "garbage in file on x86_64 platform".
   7         Root Cause: the driver registered controllers as 64-bit DMA capable
   8         for those which are not support it.
   9         Fix: Made change in the function inserting identification machanism
  10         identifying 64-bit DMA capable controllers.
  11
  12         > -----Original Message-----
  13         > From: Vasily Averin [mailto:vvs@sw.ru]
  14         > Sent: Thursday, May 04, 2006 2:49 PM
  15         > To: linux-scsi@vger.kernel.org; Kolli, Neela; Mukker, Atul;
  16         > Ju, Seokmann; Bagalkote, Sreenivas;
  17         > James.Bottomley@SteelEye.com; devel@openvz.org
  18         > Subject: megaraid_mbox: garbage in file
  19         >
  20         > Hello all,
  21         >
  22         > I've investigated customers claim on the unstable work of
  23         > their node and found a
  24         > strange effect: reading from some files leads to the
  25         >  "attempt to access beyond end of device" messages.
  26         >
  27         > I've checked filesystem, memory on the node, motherboard BIOS
  28         > version, but it
  29         > does not help and issue still has been reproduced by simple
  30         > file reading.
  31         >
  32         > Reproducer is simple:
  33         >
  34         > echo 0xffffffff >/proc/sys/dev/scsi/logging_level ;
  35         > cat /vz/private/101/root/etc/ld.so.cache >/tmp/ttt  ;
  36         > echo 0 >/proc/sys/dev/scsi/logging
  37         >
  38         > It leads to the following messages in dmesg
  39         >
  40         > sd_init_command: disk=sda, block=871769260, count=26
  41         > sda : block=871769260
  42         > sda : reading 26/26 512 byte blocks.
  43         > scsi_add_timer: scmd: f79ed980, time: 7500, (c02b1420)
  44         > sd 0:1:0:0: send 0xf79ed980                  sd 0:1:0:0:
  45         >         command: Read (10): 28 00 33 f6 24 ac 00 00 1a 00
  46         > buffer = 0xf7cfb540, bufflen = 13312, done = 0xc0366b40,
  47         > queuecommand 0xc0344010
  48         > leaving scsi_dispatch_cmnd()
  49         > scsi_delete_timer: scmd: f79ed980, rtn: 1
  50         > sd 0:1:0:0: done 0xf79ed980 SUCCESS        0 sd 0:1:0:0:
  51         >         command: Read (10): 28 00 33 f6 24 ac 00 00 1a 00
  52         > scsi host busy 1 failed 0
  53         > sd 0:1:0:0: Notifying upper driver of completion (result 0)
  54         > sd_rw_intr: sda: res=0x0
  55         > 26 sectors total, 13312 bytes done.
  56         > use_sg is 4
  57         > attempt to access beyond end of device
  58         > sda6: rw=0, want=1044134458, limit=951401367
  59         > Buffer I/O error on device sda6, logical block 522067228
  60         > attempt to access beyond end of device
  61
  62 2.      When INQUIRY with EVPD bit set issued to the MegaRAID controller,
  63         system memory gets corrupted.
  64         Root Cause: MegaRAID F/W handle the INQUIRY with EVPD bit set
  65         incorrectly.
  66         Fix: MegaRAID F/W has fixed the problem and being process of release,
  67         soon. Meanwhile, driver will filter out the request.
  68
  69 3.      One of member in the data structure of the driver leads unaligne
  70         issue on 64-bit platform.
  71         Customer reporeted "kernel unaligned access addrss" issue when
  72         application communicates with MegaRAID HBA driver.
  73         Root Cause: in uioc_t structure, one of member had misaligned and it
  74         led system to display the error message.
  75         Fix: A patch submitted to community from following folk.
  76
  77         > -----Original Message-----
  78         > From: linux-scsi-owner@vger.kernel.org
  79         > [mailto:linux-scsi-owner@vger.kernel.org] On Behalf Of Sakurai Hiroomi
  80         > Sent: Wednesday, July 12, 2006 4:20 AM
  81         > To: linux-scsi@vger.kernel.org; linux-kernel@vger.kernel.org
  82         > Subject: Re: Help: strange messages from kernel on IA64 platform
  83         >
  84         > Hi,
  85         >
  86         > I saw same message.
  87         >
  88         > When GAM(Global Array Manager) is started, The following
  89         > message output.
  90         > kernel: kernel unaligned access to 0xe0000001fe1080d4,
  91         > ip=0xa000000200053371
  92         >
  93         > The uioc structure used by ioctl is defined by packed,
  94         > the allignment of each member are disturbed.
  95         > In a 64 bit structure, the allignment of member doesn't fit 64 bit
  96         > boundary. this causes this messages.
  97         > In a 32 bit structure, we don't see the message because the allinment
  98         > of member fit 32 bit boundary even if packed is specified.
  99         >
 100         > patch
 101         > I Add 32 bit dummy member to fit 64 bit boundary. I tested.
 102         > We confirmed this patch fix the problem by IA64 server.
 103         >
 104         > **************************************************************
 105         > ****************
 106         > --- linux-2.6.9/drivers/scsi/megaraid/megaraid_ioctl.h.orig
 107         > 2006-04-03 17:13:03.000000000 +0900
 108         > +++ linux-2.6.9/drivers/scsi/megaraid/megaraid_ioctl.h
 109         > 2006-04-03 17:14:09.000000000 +0900
 110         > @@ -132,6 +132,10 @@
 111         >  /* Driver Data: */
 112         >          void __user *           user_data;
 113         >          uint32_t                user_data_len;
 114         > +
 115         > +        /* 64bit alignment */
 116         > +        uint32_t                pad_0xBC;
 117         > +
 118         >          mraid_passthru_t        __user *user_pthru;
 119         >
 120         >          mraid_passthru_t        *pthru32;
 121         > **************************************************************
 122         > ****************
 123
 124 Release Date    : Mon Apr 11 12:27:22 EST 2006 - Seokmann Ju <sju@lsil.com>
 125 Current Version : 2.20.4.8 (scsi module), 2.20.2.6 (cmm module)
 126 Older Version   : 2.20.4.7 (scsi module), 2.20.2.6 (cmm module)
 127
 128 1.      Fixed a bug in megaraid_reset_handler().
 129         Customer reported "Unable to handle kernel NULL pointer dereference
 130         at virtual address 00000000" when system goes to reset condition
 131         for some reason. It happened randomly.
 132         Root Cause: in the megaraid_reset_handler(), there is possibility not
 133         returning pending packets in the pend_list if there are multiple
 134         pending packets.
 135         Fix: Made the change in the driver so that it will return all packets
 136         in the pend_list.
 137
 138 2.      Added change request.
 139         As found in the following URL, rmb() only didn't help the
 140         problem. I had to increase the loop counter to 0xFFFFFF. (6 F's)
 141         http://marc.theaimsgroup.com/?l=linux-scsi&m=110971060502497&w=2
 142
 143         I attached a patch for your reference, too.
 144         Could you check and get this fix in your driver?
 145
 146         Best Regards,
 147         Jun'ichi Nomura
 148
 149 Release Date    : Fri Nov 11 12:27:22 EST 2005 - Seokmann Ju <sju@lsil.com>
 150 Current Version : 2.20.4.7 (scsi module), 2.20.2.6 (cmm module)
 151 Older Version   : 2.20.4.6 (scsi module), 2.20.2.6 (cmm module)
 152
 153 1.      Sorted out PCI IDs to remove megaraid support overlaps.
 154         Based on the patch from Daniel, sorted out PCI IDs along with
 155         charactor node name change from 'megadev' to 'megadev_legacy' to avoid
 156         conflict.
 157         ---
 158         Hopefully we'll be getting the build restriction zapped much sooner,
 159         but we should also be thinking about totally removing the hardware
 160         support overlap in the megaraid drivers.
 161
 162         This patch pencils in a date of Feb 06 for this, and performs some
 163         printk abuse in hope that existing legacy users might pick up on what's
 164         going on.
 165
 166         Signed-off-by: Daniel Drake <dsd@gentoo.org>
 167         ---
 168
 169 2.      Fixed a issue: megaraid always fails to reset handler.
 170         ---
 171         I found that the megaraid driver always fails to reset the
 172         adapter with the following message:
 173                 megaraid: resetting the host...
 174                 megaraid mbox: reset sequence completed successfully
 175                 megaraid: fast sync command timed out
 176                 megaraid: reservation reset failed
 177         when the "Cluster mode" of the adapter BIOS is enabled.
 178         So, whenever the reset occurs, the adapter goes to
 179         offline and just become unavailable.
 180
 181         Jun'ichi Nomura [mailto:jnomura@mtc.biglobe.ne.jp]
 182         ---
 183
 184 Release Date    : Mon Mar 07 12:27:22 EST 2005 - Seokmann Ju <sju@lsil.com>
 185 Current Version : 2.20.4.6 (scsi module), 2.20.2.6 (cmm module)
 186 Older Version   : 2.20.4.5 (scsi module), 2.20.2.5 (cmm module)
 187
 188 1.      Added IOCTL backward compatibility.
 189         Convert megaraid_mm driver to new compat_ioctl entry points.
 190         I don't have easy access to hardware, so only compile tested.
 191                 - Signed-off-by:Andi Kleen <ak@muc.de>
 192
 193 2.      megaraid_mbox fix: wrong order of arguments in memset()
 194         That, BTW, shows why cross-builds are useful-the only indication of
 195         problem had been a new warning showing up in sparse output on alpha
 196         build (number of exceeding 256 got truncated).
 197                 - Signed-off-by: Al Viro
 198                 <viro@parcelfarce.linux.theplanet.co.uk>
 199
 200 3.      Convert pci_module_init to pci_register_driver
 201         Convert from pci_module_init to pci_register_driver
 202         (from:http://kerneljanitors.org/TODO)
 203                 - Signed-off-by: Domen Puncer <domen@coderock.org>
 204
 205 4.      Use the pre defined DMA mask constants from dma-mapping.h
 206         Use the DMA_{64,32}BIT_MASK constants from dma-mapping.h when calling
 207         pci_set_dma_mask() or pci_set_consistend_dma_mask(). See
 208         http://marc.theaimsgroup.com/?t=108001993000001&r=1&w=2 for more
 209         details.
 210                 Signed-off-by: Tobias Klauser <tklauser@nuerscht.ch>
 211                 Signed-off-by: Domen Puncer <domen@coderock.org>
 212
 213 5.      Remove SSID checking for Dobson, Lindsay, and Verde based products.
 214         Checking the SSVID/SSID for controllers which have Dobson, Lindsay,
 215         and Verde is unnecessary because device ID has been assigned by LSI
 216         and it is unique value. So, all controllers with these IOPs have to be
 217         supported by the driver regardless SSVID/SSID.
 218
 219 6.      Date Thu, 27 Jan 2005 04:31:09 +0100
 220         From Herbert Poetzl <>
 221         Subject RFC: assert_spin_locked() for 2.6
 222
 223         Greetings!
 224
 225         overcautious programming will kill your kernel ;)
 226         ever thought about checking a spin_lock or even
 227         asserting that it must be held (maybe just for
 228         spinlock debugging?) ...
 229
 230         there are several checks present in the kernel
 231         where somebody does a variation on the following:
 232
 233           BUG_ON(!spin_is_locked(&some_lock));
 234
 235         so what's wrong about that? nothing, unless you
 236         compile the code with CONFIG_DEBUG_SPINLOCK but
 237         without CONFIG_SMP ... in which case the BUG()
 238         will kill your kernel ...
 239
 240         maybe it's not advised to make such assertions,
 241         but here is a solution which works for me ...
 242         (compile tested for sh, x86_64 and x86, boot/run
 243         tested for x86 only)
 244
 245         best,
 246         Herbert
 247
 248                 - Herbert Poetzl <herbert@13thfloor.at>, Thu, 27 Jan 2005
 249
 250 Release Date    : Thu Feb 03 12:27:22 EST 2005 - Seokmann Ju <sju@lsil.com>
 251 Current Version : 2.20.4.5 (scsi module), 2.20.2.5 (cmm module)
 252 Older Version   : 2.20.4.4 (scsi module), 2.20.2.4 (cmm module)
 253
 254 1.      Modified name of two attributes in scsi_host_template.
 255         On Wed, 2005-02-02 at 10:56 -0500, Ju, Seokmann wrote:
 256         > +     .sdev_attrs                     = megaraid_device_attrs,
 257         > +     .shost_attrs                    = megaraid_class_device_attrs,
 258
 259         These are, perhaps, slightly confusing names.
 260         The terms device and class_device have well defined meanings in the
 261         generic device model, neither of which is what you mean here.
 262         Why not simply megaraid_sdev_attrs and megaraid_shost_attrs?
 263
 264         Other than this, it looks fine to me too.
 265
 266 Release Date    : Thu Jan 27 00:01:03 EST 2005 - Atul Mukker <atulm@lsil.com>
 267 Current Version : 2.20.4.4 (scsi module), 2.20.2.5 (cmm module)
 268 Older Version   : 2.20.4.3 (scsi module), 2.20.2.4 (cmm module)
 269
 270 1.      Bump up the version of scsi module due to its conflict.
 271
 272 Release Date    : Thu Jan 21 00:01:03 EST 2005 - Atul Mukker <atulm@lsil.com>
 273 Current Version : 2.20.4.3 (scsi module), 2.20.2.5 (cmm module)
 274 Older Version   : 2.20.4.2 (scsi module), 2.20.2.4 (cmm module)
 275
 276 1.      Remove driver ioctl for logical drive to scsi address translation and
 277         replace with the sysfs attribute. To remove drives and change
 278         capacity, application shall now use the device attribute to get the
 279         logical drive number for a scsi device. For adding newly created
 280         logical drives, class device attribute would be required to uniquely
 281         identify each controller.
 282                 - Atul Mukker <atulm@lsil.com>
 283
 284         "James, I've been thinking about this a little more, and you may be on
 285         to something here. Let each driver add files as such:"
 286
 287                 - Matt Domsch <Matt_Domsch@dell.com>, 12.15.2004
 288                  linux-scsi mailing list
 289
 290
 291         "Then, if you simply publish your LD number as an extra parameter of
 292         the device, you can look through /sys to find it."
 293
 294                 - James Bottomley <James.Bottomley@SteelEye.com>, 01.03.2005
 295                  linux-scsi mailing list
 296
 297
 298         "I don't see why not ... it's your driver, you can publish whatever
 299         extra information you need as scsi_device attributes; that was one of
 300         the designs of the extensible attribute system."
 301
 302                 - James Bottomley <James.Bottomley@SteelEye.com>, 01.06.2005
 303                  linux-scsi mailing list
 304
 305 2.      Add AMI megaraid support - Brian King <brking@charter.net>
 306                 PCI_VENDOR_ID_AMI, PCI_DEVICE_ID_AMI_MEGARAID3,
 307                 PCI_VENDOR_ID_AMI, PCI_SUBSYS_ID_PERC3_DC,
 308
 309 3.      Make some code static - Adrian Bunk <bunk@stusta.de>
 310         Date:   Mon, 15 Nov 2004 03:14:57 +0100
 311
 312         The patch below makes some needlessly global code static.
 313         -wait_queue_head_t wait_q;
 314         +static wait_queue_head_t wait_q;
 315
 316         Signed-off-by: Adrian Bunk <bunk@stusta.de>
 317
 318 4.      Added NEC ROMB support - NEC MegaRAID PCI Express ROMB controller
 319                 PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_MEGARAID_NEC_ROMB_2E,
 320                 PCI_SUBSYS_ID_NEC, PCI_SUBSYS_ID_MEGARAID_NEC_ROMB_2E,
 321
 322 5.      Fixed Tape drive issue : For any Direct CDB command to physical device
 323         including tape, timeout value set by driver was 10 minutes. With this
 324         value, most of command will return within timeout. However, for those
 325         command like ERASE or FORMAT, it takes more than an hour depends on
 326         capacity of the device and the command could be terminated before it
 327         completes.
 328         To address this issue, the 'timeout' field in the DCDB command will
 329         have NO TIMEOUT (i.e., 4) value as its timeout on DCDB command.
 330
 331
 332
 333 Release Date    : Thu Dec  9 19:10:23 EST 2004
 334         - Sreenivas Bagalkote <sreenib@lsil.com>
 335
 336 Current Version : 2.20.4.2 (scsi module), 2.20.2.4 (cmm module)
 337 Older Version   : 2.20.4.1 (scsi module), 2.20.2.3 (cmm module)
 338
 339 i.      Introduced driver ioctl that returns scsi address for a given ld.
 340
 341         "Why can't the existing sysfs interfaces be used to do this?"
 342                 - Brian King (brking@us.ibm.com)
 343
 344         "I've looked into solving this another way, but I cannot see how
 345         to get this driver-private mapping of logical drive number-> HCTL
 346         without putting code something like this into the driver."
 347
 348         "...and by providing a mapping a function to userspace, the driver
 349         is free to change its mapping algorithm in the future if necessary .."
 350                 - Matt Domsch (Matt_Domsch@dell.com)
 351
 352 Release Date    : Thu Dec  9 19:02:14 EST 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
 353
 354 Current Version : 2.20.4.1 (scsi module), 2.20.2.3 (cmm module)
 355 Older Version   : 2.20.4.1 (scsi module), 2.20.2.2 (cmm module)
 356
 357 i.      Fix a bug in kioc's dma buffer deallocation
 358
 359 Release Date    : Thu Nov  4 18:24:56 EST 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
 360
 361 Current Version : 2.20.4.1 (scsi module), 2.20.2.2 (cmm module)
 362 Older Version   : 2.20.4.0 (scsi module), 2.20.2.1 (cmm module)
 363
 364 i.      Handle IOCTL cmd timeouts more properly.
 365
 366 ii.     pci_dma_sync_{sg,single}_for_cpu was introduced into megaraid_mbox
 367         incorrectly (instead of _for_device). Changed to appropriate
 368         pci_dma_sync_{sg,single}_for_device.
 369
 370 Release Date    : Wed Oct 06 11:15:29 EDT 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
 371 Current Version : 2.20.4.0 (scsi module), 2.20.2.1 (cmm module)
 372 Older Version   : 2.20.4.0 (scsi module), 2.20.2.0 (cmm module)
 373
 374 i.      Remove CONFIG_COMPAT around register_ioctl32_conversion
 375
 376 Release Date    : Mon Sep 27 22:15:07 EDT 2004 - Atul Mukker <atulm@lsil.com>
 377 Current Version : 2.20.4.0 (scsi module), 2.20.2.0 (cmm module)
 378 Older Version   : 2.20.3.1 (scsi module), 2.20.2.0 (cmm module)
 379
 380 i.      Fix data corruption. Because of a typo in the driver, the IO packets
 381         were wrongly shared by the ioctl path. This causes a whole IO command
 382         to be replaced by an incoming ioctl command.
 383
 384 Release Date    : Tue Aug 24 09:43:35 EDT 2004 - Atul Mukker <atulm@lsil.com>
 385 Current Version : 2.20.3.1 (scsi module), 2.20.2.0 (cmm module)
 386 Older Version   : 2.20.3.0 (scsi module), 2.20.2.0 (cmm module)
 387
 388 i.      Function reordering so that inline functions are defined before they
 389         are actually used. It is now mandatory for GCC 3.4.1 (current stable)
 390
 391         Declare some heavy-weight functions to be non-inlined,
 392         megaraid_mbox_build_cmd, megaraid_mbox_runpendq,
 393         megaraid_mbox_prepare_pthru, megaraid_mbox_prepare_epthru,
 394         megaraid_busywait_mbox
 395
 396                 - Andrew Morton <akpm@osdl.org>, 08.19.2004
 397                 linux-scsi mailing list
 398
 399         "Something else to clean up after inclusion: every instance of an
 400         inline function is actually rendered as a full function call, because
 401         the function is always used before it is defined.  Atul, please
 402         re-arrange the code to eliminate the need for most (all) of the
 403         function prototypes at the top of each file, and define (not just
 404         declare with a prototype) each inline function before its first use"
 405
 406                 - Matt Domsch <Matt_Domsch@dell.com>, 07.27.2004
 407                 linux-scsi mailing list
 408
 409
 410 ii.     Display elapsed time (countdown) while waiting for FW to boot.
 411
 412 iii.    Module compilation reorder in Makefile so that unresolved symbols do
 413         not occur when driver is compiled non-modular.
 414
 415                 Patrick J. LoPresti <patl@users.sourceforge.net>, 8.22.2004
 416                 linux-scsi mailing list
 417
 418
 419 Release Date    : Thu Aug 19 09:58:33 EDT 2004 - Atul Mukker <atulm@lsil.com>
 420 Current Version : 2.20.3.0 (scsi module), 2.20.2.0 (cmm module)
 421 Older Version   : 2.20.2.0 (scsi module), 2.20.1.0 (cmm module)
 422
 423 i.      When copying the mailbox packets, copy only first 14 bytes (for 32-bit
 424         mailboxes) and only first 22 bytes (for 64-bit mailboxes). This is to
 425         avoid getting the stale values for busy bit. We want to set the busy
 426         bit just before issuing command to the FW.
 427
 428 ii.     In the reset handling, if the reseted command is not owned by the
 429         driver, do not (wrongly) print information for the "attached" driver
 430         packet.
 431
 432 iii.    Have extended wait when issuing command in synchronous mode. This is
 433         required for the cases where the option ROM is disabled and there is
 434         no BIOS to start the controller. The FW starts to boot after receiving
 435         the first command from the driver. The current driver has 1 second
 436         timeout for the synchronous commands, which is far less than what is
 437         actually required. We now wait up to MBOX_RESET_TIME (180 seconds) for
 438         FW boot process.
 439
 440 iv.     In megaraid_mbox_product_info, clear the mailbox contents completely
 441         before preparing the command for inquiry3. This is to ensure that the
 442         FW does not get junk values in the command.
 443
 444 v.      Do away with the redundant LSI_CONFIG_COMPAT redefinition for
 445         CONFIG_COMPAT. Replace <asm/ioctl32.h> with <linux/ioctl32.h>
 446
 447                 - James Bottomley <James.Bottomley@SteelEye.com>, 08.17.2004
 448                  linux-scsi mailing list
 449
 450 vi.     Add support for 64-bit applications. Current drivers assume only
 451         32-bit applications, even on 64-bit platforms. Use the "data" and
 452         "buffer" fields of the mimd_t structure, instead of embedded 32-bit
 453         addresses in application mailbox and passthru structures.
 454
 455 vii.    Move the function declarations for the management module from
 456         megaraid_mm.h to megaraid_mm.c
 457
 458                 - Andrew Morton <akpm@osdl.org>, 08.19.2004
 459                 linux-scsi mailing list
 460
 461 viii.   Change default values for MEGARAID_NEWGEN, MEGARAID_MM, and
 462         MEGARAID_MAILBOX to 'n' in Kconfig.megaraid
 463
 464                 - Andrew Morton <akpm@osdl.org>, 08.19.2004
 465                 linux-scsi mailing list
 466
 467 ix.     replace udelay with msleep
 468
 469 x.      Typos corrected in comments and whitespace adjustments, explicit
 470         grouping of expressions.
 471
 472
 473 Release Date    : Fri Jul 23 15:22:07 EDT 2004 - Atul Mukker <atulm@lsil.com>
 474 Current Version : 2.20.2.0 (scsi module), 2.20.1.0 (cmm module)
 475 Older Version   : 2.20.1.0 (scsi module), 2.20.0.0 (cmm module)
 476
 477 i.      Add PCI ids for Acer ROMB 2E solution
 478
 479 ii.     Add PCI ids for I4
 480
 481 iii.    Typo corrected for subsys id for megaraid sata 300-4x
 482
 483 iv.     Remove yield() while mailbox handshake in synchronous commands
 484
 485
 486         "My other main gripe is things like this:
 487
 488         +       // wait for maximum 1 second for status to post
 489         +       for (i = 0; i < 40000; i++) {
 490         +               if (mbox->numstatus != 0xFF) break;
 491         +               udelay(25); yield();
 492         +       }
 493
 494         which litter the driver.  Use of yield() in drivers is deprecated."
 495
 496                 - James Bottomley <James.Bottomley@SteelEye.com>, 07.14.2004
 497                  linux-scsi mailing list
 498
 499 v.      Remove redundant __megaraid_busywait_mbox routine
 500
 501 vi.     Fix bug in the managment module, which causes a system lockup when the
 502         IO module is loaded and then unloaded, followed by executing any
 503         management utility. The current version of management module does not
 504         handle the adapter unregister properly.
 505
 506         Specifically, it still keeps a reference to the unregistered
 507         controllers. To avoid this, the static array adapters has been
 508         replaced by a dynamic list, which gets updated every time an adapter
 509         is added or removed.
 510
 511         Also, during unregistration of the IO module, the resources are
 512         now released in the exact reverse order of the allocation time
 513         sequence.
 514
 515
 516 Release Date    : Fri Jun 25 18:58:43 EDT 2004 - Atul Mukker <atulm@lsil.com>
 517 Current Version : 2.20.1.0
 518 Older Version   : megaraid 2.20.0.1
 519
 520 i.      Stale list pointer in adapter causes kernel panic when module
 521         megaraid_mbox is unloaded
 522
 523
 524 Release Date    : Thu Jun 24 20:37:11 EDT 2004 - Atul Mukker <atulm@lsil.com>
 525 Current Version : 2.20.0.1
 526 Older Version   : megaraid 2.20.0.00
 527
 528 i.      Modules are not 'y' by default, but depend on current definition of
 529         SCSI & PCI.
 530
 531 ii.     Redundant structure mraid_driver_t removed.
 532
 533 iii.    Miscellaneous indentation and goto/label fixes.
 534                 - Christoph Hellwig <hch@infradead.org>, 06.24.2004 linux-scsi
 535
 536 iv.     scsi_host_put(), do just before completing HBA shutdown.
 537
 538
 539
 540 Release Date    : Mon Jun 21 19:53:54 EDT 2004 - Atul Mukker <atulm@lsil.com>
 541 Current Version : 2.20.0.0
 542 Older Version   : megaraid 2.20.0.rc2 and 2.00.3
 543
 544 i.      Independent module to interact with userland applications and
 545         multiplex command to low level RAID module(s).
 546
 547         "Shared code in a third module, a "library module", is an acceptable
 548         solution. modprobe automatically loads dependent modules, so users
 549         running "modprobe driver1" or "modprobe driver2" would automatically
 550         load the shared library module."
 551
 552                 - Jeff Garzik <jgarzik@pobox.com> 02.25.2004 LKML
 553
 554         "As Jeff hinted, if your userspace<->driver API is consistent between
 555         your new MPT-based RAID controllers and your existing megaraid driver,
 556         then perhaps you need a single small helper module (lsiioctl or some
 557         better name), loaded by both mptraid and megaraid automatically, which
 558         handles registering the /dev/megaraid node dynamically. In this case,
 559         both mptraid and megaraid would register with lsiioctl for each
 560         adapter discovered, and lsiioctl would essentially be a switch,
 561         redirecting userspace tool ioctls to the appropriate driver."
 562
 563                 - Matt Domsch <Matt_Domsch@dell.com> 02.25.2004 LKML
 564
 565 ii.     Remove C99 initializations from pci_device id.
 566
 567         "pci_id_table_g would be much more readable when not using C99
 568         initializers.
 569         PCI table doesn't change, there's lots of users that prefer the more
 570         readable variant.  And it's really far less and much easier to grok
 571         lines without C99 initializers."
 572
 573                 - Christoph Hellwig <hch@infradead.org>, 05.28.2004 linux-scsi
 574
 575 iii.    Many fixes as suggested by Christoph Hellwig <hch@infradead.org> on
 576         linux-scsi, 05.28.2004
 577
 578 iv.     We now support up to 32 parallel ioctl commands instead of current 1.
 579         There is a conscious effort to let memory allocation not fail for ioctl
 580         commands.
 581
 582 v.      Do away with internal memory management. Use pci_pool_(create|alloc)
 583         instead.
 584
 585 vi.     Kill tasklet when unloading the driver.
 586
 587 vii.    Do not use "host_lock', driver has fine-grain locks now to protect all
 588         data structures.
 589
 590 viii.   Optimize the build scatter-gather list routine. The callers already
 591         know the data transfer address and length.
 592
 593 ix.     Better implementation of error handling and recovery. Driver now
 594         performs extended errors recovery for instances like scsi cable pull.
 595
 596 x.      Disassociate the management commands with an overlaid scsi command.
 597         Driver now treats the management packets as special packets and has a
 598         dedicated callback routine.