1 Release Date : Thu Dec 9 19:02:14 EST 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
3 Current Version : 2.20.4.1 (scsi module), 2.20.2.3 (cmm module)
4 Older Version : 2.20.4.1 (scsi module), 2.20.2.2 (cmm module)
6 i. Fix a bug in kioc's dma buffer deallocation
8 Release Date : Thu Nov 4 18:24:56 EST 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
10 Current Version : 2.20.4.1 (scsi module), 2.20.2.2 (cmm module)
11 Older Version : 2.20.4.0 (scsi module), 2.20.2.1 (cmm module)
13 i. Handle IOCTL cmd timeouts more properly.
15 ii. pci_dma_sync_{sg,single}_for_cpu was introduced into megaraid_mbox
16 incorrectly (instead of _for_device). Changed to appropriate
17 pci_dma_sync_{sg,single}_for_device.
19 Release Date : Wed Oct 06 11:15:29 EDT 2004 - Sreenivas Bagalkote <sreenib@lsil.com>
20 Current Version : 2.20.4.0 (scsi module), 2.20.2.1 (cmm module)
21 Older Version : 2.20.4.0 (scsi module), 2.20.2.0 (cmm module)
23 i. Remove CONFIG_COMPAT around register_ioctl32_conversion
25 Release Date : Mon Sep 27 22:15:07 EDT 2004 - Atul Mukker <atulm@lsil.com>
26 Current Version : 2.20.4.0 (scsi module), 2.20.2.0 (cmm module)
27 Older Version : 2.20.3.1 (scsi module), 2.20.2.0 (cmm module)
29 i. Fix data corruption. Because of a typo in the driver, the IO packets
30 were wrongly shared by the ioctl path. This causes a whole IO command
31 to be replaced by an incoming ioctl command.
33 Release Date : Tue Aug 24 09:43:35 EDT 2004 - Atul Mukker <atulm@lsil.com>
34 Current Version : 2.20.3.1 (scsi module), 2.20.2.0 (cmm module)
35 Older Version : 2.20.3.0 (scsi module), 2.20.2.0 (cmm module)
37 i. Function reordering so that inline functions are defined before they
38 are actually used. It is now mandatory for GCC 3.4.1 (current stable)
40 Declare some heavy-weight functions to be non-inlined,
41 megaraid_mbox_build_cmd, megaraid_mbox_runpendq,
42 megaraid_mbox_prepare_pthru, megaraid_mbox_prepare_epthru,
43 megaraid_busywait_mbox
45 - Andrew Morton <akpm@osdl.org>, 08.19.2004
46 linux-scsi mailing list
48 "Something else to clean up after inclusion: every instance of an
49 inline function is actually rendered as a full function call, because
50 the function is always used before it is defined. Atul, please
51 re-arrange the code to eliminate the need for most (all) of the
52 function prototypes at the top of each file, and define (not just
53 declare with a prototype) each inline function before its first use"
55 - Matt Domsch <Matt_Domsch@dell.com>, 07.27.2004
56 linux-scsi mailing list
59 ii. Display elapsed time (countdown) while waiting for FW to boot.
61 iii. Module compilation reorder in Makefile so that unresolved symbols do
62 not occur when driver is compiled non-modular.
64 Patrick J. LoPresti <patl@users.sourceforge.net>, 8.22.2004
65 linux-scsi mailing list
68 Release Date : Thu Aug 19 09:58:33 EDT 2004 - Atul Mukker <atulm@lsil.com>
69 Current Version : 2.20.3.0 (scsi module), 2.20.2.0 (cmm module)
70 Older Version : 2.20.2.0 (scsi module), 2.20.1.0 (cmm module)
72 i. When copying the mailbox packets, copy only first 14 bytes (for 32-bit
73 mailboxes) and only first 22 bytes (for 64-bit mailboxes). This is to
74 avoid getting the stale values for busy bit. We want to set the busy
75 bit just before issuing command to the FW.
77 ii. In the reset handling, if the reseted command is not owned by the
78 driver, do not (wrongly) print information for the "attached" driver
81 iii. Have extended wait when issuing command in synchronous mode. This is
82 required for the cases where the option ROM is disabled and there is
83 no BIOS to start the controller. The FW starts to boot after receiving
84 the first command from the driver. The current driver has 1 second
85 timeout for the synchronous commands, which is far less than what is
86 actually required. We now wait up to MBOX_RESET_TIME (180 seconds) for
89 iv. In megaraid_mbox_product_info, clear the mailbox contents completely
90 before preparing the command for inquiry3. This is to ensure that the
91 FW does not get junk values in the command.
93 v. Do away with the redundant LSI_CONFIG_COMPAT redefinition for
94 CONFIG_COMPAT. Replace <asm/ioctl32.h> with <linux/ioctl32.h>
96 - James Bottomley <James.Bottomley@SteelEye.com>, 08.17.2004
97 linux-scsi mailing list
99 vi. Add support for 64-bit applications. Current drivers assume only
100 32-bit applications, even on 64-bit platforms. Use the "data" and
101 "buffer" fields of the mimd_t structure, instead of embedded 32-bit
102 addresses in application mailbox and passthru structures.
104 vii. Move the function declarations for the management module from
105 megaraid_mm.h to megaraid_mm.c
107 - Andrew Morton <akpm@osdl.org>, 08.19.2004
108 linux-scsi mailing list
110 viii. Change default values for MEGARAID_NEWGEN, MEGARAID_MM, and
111 MEGARAID_MAILBOX to 'n' in Kconfig.megaraid
113 - Andrew Morton <akpm@osdl.org>, 08.19.2004
114 linux-scsi mailing list
116 ix. replace udelay with msleep
118 x. Typos corrected in comments and whitespace adjustments, explicit
119 grouping of expressions.
122 Release Date : Fri Jul 23 15:22:07 EDT 2004 - Atul Mukker <atulm@lsil.com>
123 Current Version : 2.20.2.0 (scsi module), 2.20.1.0 (cmm module)
124 Older Version : 2.20.1.0 (scsi module), 2.20.0.0 (cmm module)
126 i. Add PCI ids for Acer ROMB 2E solution
128 ii. Add PCI ids for I4
130 iii. Typo corrected for subsys id for megaraid sata 300-4x
132 iv. Remove yield() while mailbox handshake in synchronous commands
135 "My other main gripe is things like this:
137 + // wait for maximum 1 second for status to post
138 + for (i = 0; i < 40000; i++) {
139 + if (mbox->numstatus != 0xFF) break;
140 + udelay(25); yield();
143 which litter the driver. Use of yield() in drivers is deprecated."
145 - James Bottomley <James.Bottomley@SteelEye.com>, 07.14.2004
146 linux-scsi mailing list
148 v. Remove redundant __megaraid_busywait_mbox routine
150 vi. Fix bug in the managment module, which causes a system lockup when the
151 IO module is loaded and then unloaded, followed by executing any
152 management utility. The current version of management module does not
153 handle the adapter unregister properly.
155 Specifically, it still keeps a reference to the unregistered
156 controllers. To avoid this, the static array adapters has been
157 replaced by a dynamic list, which gets updated every time an adapter
160 Also, during unregistration of the IO module, the resources are
161 now released in the exact reverse order of the allocation time
165 Release Date : Fri Jun 25 18:58:43 EDT 2004 - Atul Mukker <atulm@lsil.com>
166 Current Version : 2.20.1.0
167 Older Version : megaraid 2.20.0.1
169 i. Stale list pointer in adapter causes kernel panic when module
170 megaraid_mbox is unloaded
173 Release Date : Thu Jun 24 20:37:11 EDT 2004 - Atul Mukker <atulm@lsil.com>
174 Current Version : 2.20.0.1
175 Older Version : megaraid 2.20.0.00
177 i. Modules are not 'y' by default, but depend on current definition of
180 ii. Redundant structure mraid_driver_t removed.
182 iii. Miscellaneous indentation and goto/label fixes.
183 - Christoph Hellwig <hch@infradead.org>, 06.24.2004 linux-scsi
185 iv. scsi_host_put(), do just before completing HBA shutdown.
189 Release Date : Mon Jun 21 19:53:54 EDT 2004 - Atul Mukker <atulm@lsil.com>
190 Current Version : 2.20.0.0
191 Older Version : megaraid 2.20.0.rc2 and 2.00.3
193 i. Independent module to interact with userland applications and
194 multiplex command to low level RAID module(s).
196 "Shared code in a third module, a "library module", is an acceptable
197 solution. modprobe automatically loads dependent modules, so users
198 running "modprobe driver1" or "modprobe driver2" would automatically
199 load the shared library module."
201 - Jeff Garzik <jgarzik@pobox.com> 02.25.2004 LKML
203 "As Jeff hinted, if your userspace<->driver API is consistent between
204 your new MPT-based RAID controllers and your existing megaraid driver,
205 then perhaps you need a single small helper module (lsiioctl or some
206 better name), loaded by both mptraid and megaraid automatically, which
207 handles registering the /dev/megaraid node dynamically. In this case,
208 both mptraid and megaraid would register with lsiioctl for each
209 adapter discovered, and lsiioctl would essentially be a switch,
210 redirecting userspace tool ioctls to the appropriate driver."
212 - Matt Domsch <Matt_Domsch@dell.com> 02.25.2004 LKML
214 ii. Remove C99 initializations from pci_device id.
216 "pci_id_table_g would be much more readable when not using C99
218 PCI table doesn't change, there's lots of users that prefer the more
219 readable variant. And it's really far less and much easier to grok
220 lines without C99 initializers."
222 - Christoph Hellwig <hch@infradead.org>, 05.28.2004 linux-scsi
224 iii. Many fixes as suggested by Christoph Hellwig <hch@infradead.org> on
225 linux-scsi, 05.28.2004
227 iv. We now support up to 32 parallel ioctl commands instead of current 1.
228 There is a conscious effort to let memory allocation not fail for ioctl
231 v. Do away with internal memory management. Use pci_pool_(create|alloc)
234 vi. Kill tasklet when unloading the driver.
236 vii. Do not use "host_lock', driver has fine-grain locks now to protect all
239 viii. Optimize the build scatter-gather list routine. The callers already
240 know the data transfer address and length.
242 ix. Better implementation of error handling and recovery. Driver now
243 performs extended errors recovery for instances like scsi cable pull.
245 x. Disassociate the management commands with an overlaid scsi command.
246 Driver now treats the management packets as special packets and has a
247 dedicated callback routine.