b43: Remove unnecessary MMIO accesses in the interrupt hotpath.
[openwrt.git] / target / linux / generic-2.6 / files / crypto / ocf / ChangeLog
1 2008-09-18 01:27 davidm
2
3 * Makefile, README, README.sglinux,
4 patches/linux-2.6.26-natt.patch, patches/linux-2.6.26-ocf.patch,
5 patches/openssl-0.9.8g.patch, patches/openssl-0.9.8i.patch:
6
7 Updates for a new OCF release with openssl-0.9.8i and linux-2.6.26
8 support.
9
10 2008-09-18 00:19 davidm
11
12 * Config.in, Kconfig, Makefile, ep80579/Makefile,
13 ep80579/environment.mk, ep80579/icp_asym.c, ep80579/icp_common.c,
14 ep80579/icp_ocf.h, ep80579/icp_sym.c,
15 ep80579/linux_2.6_kernel_space.mk:
16
17 A new driver from Intel for their Intel QuickAssist enabled EP80579
18 Integrated Processor Product Line.
19
20 Adrian Hoban Brad Vrabete
21
22
23 2008-07-25 01:01 gerg
24
25 * ocf-compat.h:
26
27 From linux-2.6.26 onwards there is now a linux/fdtable.h include
28 that contains the file_fdtable() definition.
29
30 2008-07-05 01:20 davidm
31
32 * Makefile, patches/linux-2.6.25-natt.patch,
33 patches/linux-2.6.25-ocf.patch:
34
35 A new ocf-linux release and some patches to send to the OS guys.
36
37 2008-07-03 21:21 davidm
38
39 * crypto.c:
40
41 Clean up a some possible deadlock/busy wait issues with
42 locking/sleeping. This has greatly improved openswan 2.6.14
43 reliability ;-)
44
45 Make more of our state available in /sys for debugging.
46
47 2008-06-23 20:38 davidm
48
49 * Config.in, Kconfig:
50
51 Some settings should be bool's John Gumb
52
53 2008-05-28 03:43 philipc
54
55 * ixp4xx/ixp4xx.c: Revert previous checkin since the
56 callbacks are meant to complete the crypto request. Before
57 reverting this, CONFIG_DEBUG_SLAB gives: slab error in
58 verify_redzone_free(): cache `ixp4xx_q': double free detected when
59 trying to ssh to the device.
60
61 2008-04-27 19:31 davidm
62
63 * ixp4xx/ixp4xx.c:
64
65 We were not completing a crypto request under some error
66 conditions.
67
68 2008-04-02 01:51 davidm
69
70 * talitos/talitos.c:
71
72 This patch is for crypto/ocf/talitos for use on linux 2.6.23. It
73 is applied to the ocf-linux-20071215 release.
74
75 Signed-off-by: Lee Nipper
76
77 2008-02-29 00:43 davidm
78
79 * crypto.c, ocf-compat.h, ixp4xx/ixp4xx.c:
80
81 freshen up the 2.4 support, some recent OCF changes and openswan
82 changes are not that old-os friendly.
83
84 Force OCF to select HW/SW, otherwise it may get stuck on the first
85 device. This change means we will favour HW over SW, but we will
86 use both as required. Passing in a crid of 0 to crypto_newsession
87 effectively meant we were stuck on the first device registered with
88 OCF, not good. This only applied to ipsec, cryptodev already did
89 the right thing.
90
91 2008-01-31 07:37 gerg
92
93 * hifn/hifn7751.c:
94
95 The linux-2.6.24 modules build fails if the pci ID table doesn't
96 have a NULL entry at the end. So add one.
97
98 2008-01-29 09:16 gerg
99
100 * cryptosoft.c, ocf-compat.h:
101
102 Added some compatability macros for scatterlist changes from 2.6.24
103 onwards.
104
105 2007-12-16 07:31 davidm
106
107 * Makefile:
108
109 missed an openssl patch name change
110
111 2007-12-16 07:27 davidm
112
113 * Makefile, README, README.sglinux, patches/linux-2.6.23-ocf.patch,
114 patches/openssl-0.9.8e.patch, patches/openssl-0.9.8g.patch,
115 patches/ssl.patch:
116
117 updates for a new ocf release and associated bits
118
119 2007-12-16 06:36 davidm
120
121 * crypto.c:
122
123 Be very careful what you do while potentially in an driver unload
124 state or we will call through NULL pointers.
125
126 Reported by Nawang Chhetan .
127
128 2007-12-14 22:32 davidm
129
130 * cryptodev.c:
131
132 Add in an unlock_ioctl when available to help SMP systems a lot.
133 Otherwise all ioctls get a BKL :-(
134
135 Problem found by Egor N. Martovetsky
136
137 2007-12-14 18:29 davidm
138
139 * cryptosoft.c:
140
141 reformat the alg table to make it easier to read.
142
143 2007-12-14 18:29 davidm
144
145 * crypto.c:
146
147 Fix more driver locking/sleeping bugs report by Nawang Chhetan
148
149
150 2007-12-12 21:36 davidm
151
152 * Config.in, Kconfig, Makefile, pasemi/Makefile, pasemi/pasemi.c,
153 pasemi/pasemi_fnu.h:
154
155 Here I'm including my PA Semi driver patch to OCF. Please consider
156 it for inclusion into next OCF release.
157
158 Egor N. Martovetsky
159
160 2007-12-05 00:37 davidm
161
162 * patches/: linux-2.4.35-ocf.patch, linux-2.6.22-ocf.patch:
163
164 More correct count setting if we get a signal
165
166 Adrian Hoban
167
168 2007-12-05 00:02 davidm
169
170 * random.c:
171
172 OCF has a static array for holding random data. The random number
173 generator I have can write directly into physically contiguous
174 memory. Static memory comes from the heap and isn't physically
175 contiguous. I could use kmalloc'd memory and then copy into the OCF
176 static buf but I'd like to avoid a memory copy. The following patch
177 (Physically_Contig_Rand.patch) allows me to avoid a memory copy and
178 should not impact the other OCF drivers:
179
180 Adrian Hoban
181
182 2007-12-05 00:01 davidm
183
184 * Kconfig:
185
186 Fix a typo in the Kconfig
187
188 2007-11-23 19:15 davidm
189
190 * talitos/talitos_dev.h:
191
192 fix the DPRINTF macro so that it actually compiles.
193
194 2007-11-22 19:41 davidm
195
196 * ocf-compat.h, talitos/talitos.c, talitos/talitos_dev.h:
197
198 Various updates to get talitos compiling and work on real-world (ie
199 ubuntu) kernels.
200
201 2007-11-08 02:21 davidm
202
203 * crypto.c:
204
205 keep and eye on us being completely blocked. If we have Q's to
206 process, but all the requests are blocked, sleep. We do not want
207 to busy loop until a driver unblocks as it uses valuable CPU
208 resources that could be doing something much more important ;-)
209
210 2007-11-07 19:04 davidm
211
212 * hifn/hifn7751.c:
213
214 hifn driver was failing to unblock itself under some "out of
215 resources" conditions. It would return ERESTART to signal it was
216 full but never call crypto_unblock to start things moving again.
217
218 2007-11-06 02:09 davidm
219
220 * hifn/hifn7751.c:
221
222 Remove some bogus trace left in the driver for the overflow (too
223 busy) case.
224
225 2007-10-12 21:10 gerg
226
227 * crypto.c, ixp4xx/ixp4xx.c:
228
229 Fix up use of kmem_cache_create() - it takes one less argument in
230 2.6.23 onwards.
231
232 2007-10-03 02:41 gerg
233
234 * ixp4xx/Makefile:
235
236 The directory locations for includes in CSR-2.4 is different. Need
237 to modify the CFLAGS accordingly if using CSR-2.4.
238
239 2007-09-22 00:39 philipc
240
241 * ixp4xx/Makefile: linux 2.4 make dep was failing. This is
242 a quick fix to get it building, need to double check this.
243
244 2007-09-19 00:13 mmccreat
245
246 * Config.in: Add config option CONFIG_OCF_IXP4XX_SHA1_MD5,
247 that enables SHA1 and MD5 hashing to be done by the IXP4xx crypto
248 accelerator (although it is much slower than using cryptosoft).
249
250 2007-09-18 21:45 mmccreat
251
252 * Makefile, random.c: - Force the inclusion of autoconf.h,
253 which contains #defines for CONFIG_xxx options for OCF. -
254 Removing additional -D option, now that we are including the
255 CONFIG_xxx #defines.
256
257 2007-09-18 21:44 mmccreat
258
259 * Kconfig: Add config option CONFIG_OCF_IXP4XX_SHA1_MD5,
260 that enables SHA1 and MD5 hashing to be done by the IXP4xx crypto
261 accelerator (although it is much slower than using cryptosoft).
262
263 2007-09-18 21:37 mmccreat
264
265 * cryptodev.h: The CRYPTO_MAX_DATA_LEN limit should be
266 0xFFFF ie 64K - 1.
267
268 2007-09-18 21:19 mmccreat
269
270 * ixp4xx/ixp4xx.c: - Rework the code so that the correct IXP
271 function, ixCryptoAccHashPerform(), is used to calculate SHA1 and
272 MD5 hashes. NB: The performance of using the IXP4xx hardware is
273 really, really poor compared to using cryptosoft (and the kernel
274 crypto). - Only support SHA1 and MD5 hashing if the
275 CONFIG_OCF_IXP4XX_SHA1_MD5 is enabled.
276
277 2007-08-30 21:42 davidm
278
279 * Makefile:
280
281 do not archive build files in the crypto-tools archive
282
283 2007-08-22 19:19 mmccreat
284
285 * cryptodev.c: Fix up the checking for key lengths, when the
286 key can be of unlimited size.
287
288 2007-08-16 01:50 davidm
289
290 * Makefile:
291
292 Better 2.4 compat for "make dep" now working with fastdep.
293
294 2007-07-28 08:25 davidm
295
296 * Makefile, README, README.sglinux, patches/crypto-tools.patch,
297 patches/linux-2.4.29-ocf.patch, patches/linux-2.4.35-ocf.patch,
298 patches/linux-2.6.11-ocf.patch, patches/linux-2.6.22-ocf.patch,
299 patches/ssl.patch:
300
301 Update all the patches and put the patch making target back into
302 the Makefile.
303
304 2007-07-28 08:25 davidm
305
306 * hifn/hifn7751.c:
307
308 fix an unused variable warning when HARVESTING is disabled
309
310 2007-07-27 21:33 davidm
311
312 * hifn/hifn7751.c, ixp4xx/ixp4xx.c, safe/safe.c, talitos/talitos.c:
313
314
315 Remove all the random code if OCF does not have radom harvesting
316 enabled.
317
318 2007-07-26 00:36 davidm
319
320 * Kconfig, hifn/hifnHIPP.c, hifn/hifnHIPPvar.h:
321
322 Changes to get the hifn HIPP stub driver to build.
323
324 2007-07-25 21:25 davidm
325
326 * Makefile, hifn/Makefile, hifn/hifnHIPP.c, hifn/hifnHIPPreg.h,
327 hifn/hifnHIPPvar.h, ixp4xx/Makefile, ocfnull/Makefile,
328 safe/Makefile, talitos/Makefile:
329
330 Bring in the hifnHIPP driver written by Xelerance. This is the
331 super hifn chip with full protocol offload.
332
333 Switch to much more traditional Makefile/subdir building. The
334 Makefiles are nicer now, but still not beautiful, 2.6 and 2.4
335 capable builds result in a certain amount of uglyiness.
336
337 2007-07-24 21:46 davidm
338
339 * cryptodev.c:
340
341 Clean up all the driver id checking and session management so
342 adding/removing drivers all continues to run cleanly.
343
344 2007-07-24 20:14 davidm
345
346 * talitos/talitos.c:
347
348 From: Ahsan Kabir
349
350 less than .1% packet corruption was detected using the talitos
351 driver. It turns out we don't need the cipher iv out len/ptr field
352 to do ESP IPsec. Therefore we set the len field as 0, which tells
353 the SEC not to do anything with this len/ptr field.
354
355 Signed-off-by: Ahsan Kabir
356 Signed-off-by: Kim Phillips
357
358 2007-07-24 08:25 davidm
359
360 * cryptosoft.c, ocf-bench.c, hifn/hifn7751.c, ixp4xx/ixp4xx.c,
361 safe/safe.c, talitos/talitos.c:
362
363 Switch the remaining GFP_ATOMIC to the newer and not deprecated
364 SLAB_ATOMIC
365
366 2007-07-23 22:16 mmccreat
367
368 * ixp4xx/ixp4xx.c: Add missing ";" from end of dprintk()
369 call!
370
371 2007-07-21 01:16 davidm
372
373 * ocf-compat.h, rndtest.c, hifn/hifn7751.c, safe/safe.c:
374
375 pci_register_driver is nothing like it is depending on the kernel,
376 so we need a compat function to fix it up for all kernels before
377 2.6.10, and in different ways for older and not so older versions.
378
379 2007-07-20 21:54 davidm
380
381 * safe/safe.c:
382
383 Make the debug macro ';' safe so you do not get compiler warnings
384
385 2007-07-20 21:53 davidm
386
387 * talitos/: talitos.c, talitos_dev.h, talitos_soft.h:
388
389 update to the latest FreeBSD driver structure and fix up the code
390 as required.
391
392 2007-07-20 03:07 davidm
393
394 * rndtest.c:
395
396 more headers needed to compile on 2.4
397
398 2007-07-20 03:00 davidm
399
400 * cryptosoft.c:
401
402 Put in the 2.4 stubs to support compression
403
404 2007-07-20 02:53 davidm
405
406 * crypto.c, ocf-compat.h:
407
408 move some more compat stuff into the compat header.
409
410 2007-07-20 02:47 davidm
411
412 * talitos/talitos.c:
413
414 support of_platform_driver for newer, ARCH=powerpc based kernels.
415 Signed-off-by: Kim Phillips
416
417 2007-07-20 02:46 davidm
418
419 * talitos/talitos.c:
420
421 From: Ahsan Kabir
422
423 When Talitos completes job both the channel and execution unit are
424 capable of generating interrupts. Talitos used to take two
425 interrupts per request - one for channel completion and the other
426 for execution unit completion. This patch ensures that Talitos
427 takes interrupt only for channel completion. Execution unit will
428 generate interrupt only when there is error and the error
429 interrupts for execution units are not masked.
430
431 Signed-off-by: Ahsan Kabir
432 Signed-off-by: Kim Phillips
433
434 2007-07-20 02:37 davidm
435
436 * cryptodev.c:
437
438 Unless the user specifies, select from both HW and SW.
439
440 Clean up some debug to report the actual ioctl name.
441
442 Compiler warning on newer compilers.
443
444 2007-07-20 02:35 davidm
445
446 * cryptodev.h:
447
448 moved dprintk to the compat code, seemed nicer in there.
449
450 2007-07-20 02:35 davidm
451
452 * rndtest.c:
453
454 Need to clean up some warnings etc, more includes
455
456 2007-07-20 02:34 davidm
457
458 * ocf-compat.h, hifn/hifn7751.c, safe/safe.c:
459
460 new shared IRQ flags for 2.6.22 and a safer version of the debug
461 macro
462
463 2007-07-20 00:52 davidm
464
465 * cryptosoft.c:
466
467 Implement compression based on the code from the openswan guys.
468
469 2007-07-20 00:52 davidm
470
471 * criov.c:
472
473 Fix compiler warning on non-value returning void func.
474
475 2007-07-18 22:55 davidm
476
477 * hifn/hifn7751.c, safe/safe.c:
478
479 Use pci_register_driver rather than pci_module_init.
480 pci_module_init has been dropped in 2.6.22 yet pci_register_driver
481 has always existed and used to do some crazy hotplug junk.
482
483 2007-07-18 21:55 gerg
484
485 * ixp4xx/ixp4xx.c:
486
487 Added a missing ";" at the end of the ixp_kproces() prototype. It
488 is IXP465 specific, only showed up when generateing for SG720.
489
490 2007-07-17 00:37 davidm
491
492 * Makefile, cryptodev.c, random.c, rndtest.c, hifn/hifn7751.c,
493 ocfnull/ocfnull.c, safe/safe.c:
494
495 Fixup all the debug support for 2.4 kernels, clean up the entropy
496 harvester to be far more robust.
497
498 2007-07-14 02:19 davidm
499
500 * talitos/talitos.c:
501
502 Old patch that had not been applied Kim Phillips
503
504
505 2007-07-14 01:12 davidm
506
507 * Config.in, Kconfig, Makefile, cryptodev.c, cryptodev.h, random.c,
508 syscall.h:
509
510 Finally ditched all the syscall stuff. You can now enable/disable
511 the random harvestor. Pulled in most of random.c from openswan
512 project and fixed some obvious bugs (that were always there).
513
514 2007-07-13 21:59 davidm
515
516 * ocf-compat.h:
517
518 Better error printing and checking for drivers
519
520 2007-07-13 21:56 davidm
521
522 * cryptosoft.c:
523
524 Fix some incorrect debug (reporting wrong error type)
525
526 2007-07-13 21:55 davidm
527
528 * hifn/hifn7751.c, safe/safe.c:
529
530 Make the code more similar to Free-BSD by reverting to the same
531 debug macros
532
533 2007-07-13 21:53 davidm
534
535 * ocfnull/ocfnull.c:
536
537 Fix up the null driver to work again in the new framework.
538
539 2007-07-06 23:54 mmccreat
540
541 * cryptodev.c, cryptodev.h, cryptosoft.c: - Update OCF to
542 work with new Crypto API introduced in 2.6.19 kerneli, and add
543 macros so it work with older kernels. - Add support for SHA256,
544 SHA384 and SHA512 HASH and HMAC algorithms. - Cryptosoft: Only
545 register algorithms that the kernel has implementations for.
546
547 2007-07-03 19:52 davidm
548
549 * Kconfig, README, README.sglinux, criov.c, crypto.c, cryptodev.c,
550 cryptodev.h, cryptosoft.c, ocf-bench.c, ocf-compat.h, random.c,
551 rndtest.c, uio.h, hifn/hifn7751.c, hifn/hifn7751reg.h,
552 hifn/hifn7751var.h, ixp4xx/ixp4xx.c, ocfnull/ocfnull.c,
553 safe/safe.c, safe/safevar.h, talitos/talitos.c:
554
555 Updated OCF to the lastest FreeBSD version.
556
557 There was a lot of change in here, some of which will help FIP's,
558 some which won't.
559
560 Did lots of cleaning and diff reduction against the freebsd code.
561 Still more cleaning to do.
562
563 2007-06-01 21:58 gerg
564
565 * Config.in:
566
567 Put the regular old Config.in back, needed for puclic releases.
568
569 2007-06-01 21:58 gerg
570
571 * Kconfig:
572
573 Change the CONFIG_OCF_IXP400 dependencies to be the same as they
574 where in the Config.in file.
575
576 2007-05-28 21:40 gerg
577
578 * Config.in, Kconfig:
579
580 Switch all module configuration over to new style Kconfigs.
581
582 2007-05-24 18:49 davidm
583
584 * cryptodev.c, random.c:
585
586 Work around some problems on redhat systems with errno redefinition
587
588 2007-04-30 21:09 gerg
589
590 * cryptosoft.c:
591
592 The CRYPTO_TFM_MODE_ family of defines no longer exists from 2.6.21
593 onwards. As far as I can tell you don't need to pass it to the
594 crypto_alloc_tfm() function anymore.
595
596 So define it to be 0 if it doesn't exist.
597
598 2007-04-03 02:13 gerg
599
600 * syscall.h:
601
602 Added syscall macros for SH architecture. Just temporary, 'till
603 Dave fixes the OCF code to not use syscalls from the modules :-)
604
605 2007-02-16 23:10 davidm
606
607 * syscall.h:
608
609 ensure the temprary 2.6 fix doesn't break 2.4
610
611 2007-02-07 22:23 gerg
612
613 * cryptodev.c, random.c, syscall.h:
614
615 Temporary fix for new 2.6 kernels no longer defining in-kernel
616 system call functions. Define them locally for now until we fix
617 properly.
618
619 2007-02-07 03:10 gerg
620
621 * ixp4xx/ixp4xx.c:
622
623 Changes to support the different INIT_WORK() mechanism from kernels
624 2.6.20 onwards.
625
626 2007-02-06 02:38 gerg
627
628 * crypto.c:
629
630 Cleaned up use of kmem_cache_t and use of SLAB_ATOMIC.
631
632 2006-12-05 20:50 cpascoe
633
634 * hifn/hifn7751.c, safe/safe.c, talitos/talitos.c:
635
636 Remove pt_regs from OCF interrupt handlers for 2.6.19+
637
638 2006-12-02 03:36 gerg
639
640 * criov.c, crypto.c, cryptodev.c, cryptosoft.c, ocf-bench.c,
641 random.c, rndtest.c, hifn/hifn7751.c, ixp4xx/ixp4xx.c,
642 ocfnull/ocfnull.c, safe/safe.c, talitos/talitos.c:
643
644 Can no longer include linux/config.h as of 2.6.19 kernels. Need to
645 conditionally include it based on AUTOCONF_INCLUDED not being
646 defined.
647
648 2006-10-13 21:52 cpascoe
649
650 * random.c:
651
652 Remove another race condition that may result in us running more
653 than one random thread if modules are reloaded during heavy system
654 load.
655
656 2006-10-13 21:18 cpascoe
657
658 * random.c:
659
660 - Permit delivery of SIGKILL to the random thread. - Do not exit
661 prematurely if poll() is interrupted. - Improve exit conditions so
662 that we quit immediately, rather than loop infinitely, if the
663 last RNG source is removed while we are trying to fill our
664 buffer.
665
666 2006-10-03 20:28 cpascoe
667
668 * crypto.c, crypto.c: AutoMerged
669 >
670 > Zero the empty half of "new driver" buffer, and not past the end
671 of the old
672 > buffer that we are about to free.
673
674 2006-10-03 20:28 cpascoe
675
676 * crypto.c:
677
678 Zero the empty half of "new driver" buffer, and not past the end of
679 the old buffer that we are about to free.
680
681 2006-08-25 23:57 davidm
682
683 * cryptosoft.c:
684
685 Do not print errors for failed ALG requests unless debugging
686
687 2006-07-14 21:44 davidm
688
689 * cryptodev.h:
690
691 2.6.11 and earlier did not have the files_fdtable macro.
692
693 2006-06-21 21:26 gerg
694
695 * cryptodev.h, hifn/hifn7751.c, ocfnull/ocfnull.c, safe/safe.c,
696 talitos/talitos.c:
697
698 Fixed up more occurrances of MODULE_PARM() needing to be converted
699 to module_param() (as of 2.6.17 and onwards).
700
701 2006-06-21 00:28 gerg
702
703 * cryptosoft.c:
704
705 Change use of MODULE_PARM to module_param() for 2.6.17+ kernels.
706
707 2006-06-20 22:13 gerg
708
709 * crypto.c, cryptodev.c, cryptodev.h, ocf-bench.c, ixp4xx/ixp4xx.c:
710
711
712 As of 2.6.17 and onwards module_param is used in place of
713 MODULE_PARM.
714
715 2006-06-06 00:31 gerg
716
717 * Makefile:
718
719 Fix Makefile to find includes is using CSR-2.1.
720
721 2006-05-31 01:44 gerg
722
723 * cryptodev.h:
724
725 The vars "crypto_usercrypto", "crypto_userasymcrypto" are declared
726 as extern in the header file, but static in the c file. I guessed
727 that they should probably be truely static, so removed the extern
728 declarations from the header.
729
730 2006-05-25 21:06 davidm
731
732 * talitos/talitos.c:
733
734 This fixes a situation that I never provably experienced, where a
735 descriptor in memory may be reserved within the proper lock, and
736 freed immediately after, only for a few cycles, right outside the
737 lock. Kim Phillips
738
739 2006-05-15 19:49 davidm
740
741 * criov.c, crypto.c, cryptodev.c, cryptodev.h, cryptosoft.c,
742 ocf-bench.c, random.c, rndtest.c, uio.h, ixp4xx/ixp4xx.c:
743
744 Remove the "all rights reserved" from the Intel copyrights.
745
746 2006-05-12 21:19 davidm
747
748 * Config.in, Kconfig, Makefile, cryptodev.c, ocf-bench.c,
749 ocfnull/ocfnull.c:
750
751 Add in a null OCF driver that does nothing at all, useful for
752 measuring the cost of various parts of the ipsec stack.
753
754 2006-05-12 21:17 davidm
755
756 * ixp4xx/ixp4xx.c:
757
758 make sure we do not overwrite a correctly set error type.
759
760 2006-05-12 06:52 davidm
761
762 * crypto.c:
763
764 Fix a problem where a driver would return ERESTART (full) but then
765 unblock itself before the upper layer had marked it as blocked.
766 This caused the code to get stuck in crypto_proc and process no
767 more requests.
768
769 2006-05-12 06:47 davidm
770
771 * cryptosoft.c:
772
773 Implement CRD_F_KEY_EXPLICIT for cryptosoft so keys can be changed
774 on an active session.
775
776 2006-05-10 20:09 davidm
777
778 * README, criov.c, crypto.c, cryptodev.h, cryptosoft.c,
779 ocf-bench.c, random.c, rndtest.c, uio.h, hifn/hifn7751.c,
780 ixp4xx/ixp4xx.c, safe/safe.c, safe/safevar.h:
781
782 update email addresses and other house cleaning
783
784 2006-05-10 20:08 davidm
785
786 * cryptodev.c:
787
788 pull in better error checking from openswan modifications
789
790 2006-05-10 19:11 davidm
791
792 * cryptosoft.c:
793
794 Fix an unused variable warning when various options are disabled.
795
796 2006-05-10 19:10 davidm
797
798 * cryptodev.h:
799
800 Add support for 2.4 kernels for the new FD cloning operation
801
802 2006-05-09 19:48 davidm
803
804 * hifn/hifn7751.c:
805
806 remove the hifn 7855 support, this driver will never work with
807 that chip.
808
809 2006-05-08 23:34 davidm
810
811 * hifn/hifn7751var.h:
812
813 Contiguous buffer support so that ocf-bench can run properly.
814
815 2006-05-05 23:21 davidm
816
817 * hifn/hifn7751.c:
818
819 Add in contiguous buffer support so that ocf-bench and run on the
820 driver.
821
822 2006-05-05 23:14 davidm
823
824 * ocf-bench.c:
825
826 Our requests were out of order, need to do crypto then auth on
827 encrypt requests. Some drivers enforce this.
828
829 2006-05-04 23:21 davidm
830
831 * crypto.c:
832
833 Do not run "newsession" with lock, since newsession may sleep on
834 some targets. Handle the accounting so that things are not pulled
835 from underneath us.
836
837 2006-05-04 23:20 davidm
838
839 * cryptodev.c:
840
841 Switch to a less optimal (marginally) solution for creating a new
842 fd that appears to work in far more versions of the kernel
843 including 64bit versions.
844
845 2006-05-04 18:54 davidm
846
847 * ocf-bench.c:
848
849 Turn off the IXP access lib benchmarking by default as most people
850 don't have it.
851
852 Paul Wouters
853
854 2006-04-01 08:23 davidm
855
856 * Makefile:
857
858 Remove more temp files when cleaning
859
860 2006-04-01 08:12 davidm
861
862 * hifn/hifn7751reg.h:
863
864 7855 PCI id's as yet untested
865
866 2006-04-01 08:08 davidm
867
868 * hifn/hifn7751.c:
869
870 add PCI id's for the 7855 and AES support, card is untested still
871 as it requires 128MB of PCI memory !
872
873 2006-03-31 08:38 davidm
874
875 * README.sglinux:
876
877 small update to instructions with corrected patch name
878
879 2006-03-31 00:23 davidm
880
881 * Config.in:
882
883 Add the Talitos driver to the 2.4 config, even though it probably
884 won't compile.
885
886 2006-03-30 07:48 davidm
887
888 * Kconfig, Makefile, talitos/talitos.c, talitos/talitos_dev.h,
889 talitos/talitos_soft.h:
890
891 Please find attached the freescale SEC driver for OCF. It's been
892 (most recently) tested on an SEC2.0 based MPC8541E
893 (cryptographically identical to the MPC8555E) under 2.6.15.2, with
894 openssl-0.9.8a and openswan-2.4.3 (2.4.5rc5 won't keep the security
895 association up for me for some reason).
896
897 Please feel free to add it to your next release of OCF-Linux :-)
898
899 Kim Phillips
900
901 2006-03-20 19:34 davidm
902
903 * safe/: safe.c, safevar.h:
904
905 Safenet 1141 v1.0 chips have a DMA lockup error if you access the
906 chip while DMA'ing. As a work around you need to limit your DMA to
907 256 byte chunks using scatter/gather descriptors. All the SG/SME
908 products have v1.0 chips and would lockup with more than two
909 outstanding packets at one time.
910
911 Fix the KASSERT macro
912
913 Add some more exhaustive initialisation.
914
915 2006-03-15 21:58 davidm
916
917 * cryptodev.h, random.c:
918
919 Switch random support to "ints" since that is what the kernel uses
920 and not using the same thing is 64bit wise a bad idea.
921
922 Fix FIP's code to ensure correct amount of data is passed in.
923
924 Add work around for broken 64bit OS RNG support (disable it)
925
926 General code cleanups.
927
928 2006-03-15 21:55 davidm
929
930 * hifn/hifn7751.c:
931
932 Fixes for 64bit OS's, make sure PCI address are within bus space,
933 make sure we order writes to the bus so that chip functions
934 correctly. Some small cleanups.
935
936 2006-03-15 21:48 davidm
937
938 * hifn/hifn7751var.h:
939
940 Remove unused field from structure
941
942 2006-03-15 21:47 davidm
943
944 * safe/safe.c:
945
946 Make the code more 64bit OS compatible, force PCI address space
947 and so on.
948
949 2006-03-09 20:42 davidm
950
951 * ixp4xx/ixp4xx.c:
952
953 If we call ixpCryptoAccInit() and it fails, just assume that it
954 has already been called. This allows our "rc" scripts to be
955 openswan and freeswan compatible ore easily.
956
957 2006-02-28 23:21 davidm
958
959 * README:
960
961 generalise it a bit so rel-dates don't get in the way
962
963 2006-02-28 01:52 davidm
964
965 * README, patches/ssh.patch, patches/ssl.patch:
966
967 Updated the README and patches for a release
968
969 2006-02-25 09:21 davidm
970
971 * README, README.sglinux:
972
973 updates for a new release of OCF, ssl patches and so on.
974
975 2006-02-25 08:44 davidm
976
977 * crypto.c:
978
979 We were calling the "process"routines with interrupts disabled. I
980 can see no good reason for this and it provokes badness warnings in
981 2.6 under some conditions.
982
983 I am going to run with the Q's unlocked for processing, and
984 hopefully it will allow the system to be more responsive. It
985 hasn't affected ipsec throughput in any way. Userland throughput
986 (multi threaded) may have improved significantly though, but it
987 needs more testing.
988
989 2006-02-24 23:32 davidm
990
991 * cryptodev.c:
992
993 Whoa, set the segments with uninitted values can't be good. Clean
994 out the rest of the old code that was accidently left in.
995
996 2006-02-22 01:02 davidm
997
998 * cryptodev.c:
999
1000 Still not sure about this one, but this is working for all the
1001 cases I can see so far. If it gets weird, I am going to switch to
1002 a simple clone and chain the sessions rather than a new fcr per fd.
1003
1004 2006-02-20 22:12 davidm
1005
1006 * cryptodev.c:
1007
1008 Error handling case could free data that was not allocated
1009
1010 Ronen Shitrit Marvell Semiconductor Israel
1011 Ltd
1012
1013 2006-02-20 21:57 davidm
1014
1015 * cryptosoft.c:
1016
1017 Proper SHA/MD5 (non hmac) implementation, remove some retrictions
1018 for hashes (which only applied to crypto) and fix over zealous
1019 error checking.
1020
1021 Ronen Shitrit Marvell Semiconductor Israel
1022 Ltd
1023
1024 2006-02-09 21:15 davidm
1025
1026 * cryptodev.c:
1027
1028 The code wasn't quite right and needed some fixing for proper file
1029 accounting.
1030
1031 2006-01-24 20:08 davidm
1032
1033 * crypto.c, cryptodev.c, hifn/hifn7751.c:
1034
1035 We implement our own open in the CRIOGET ioctl now which does
1036 pretty much the same thing as the BSD version, rather than use an
1037 open system call which gets caught out by chroot.
1038
1039 2006-01-06 00:42 gerg
1040
1041 * cryptosoft.c:
1042
1043 Moved "{" block marker inside the set of
1044 "defined(CONFIG_CRYPTO_HMAC) || defined(CONFIG_CRYPTO_HMAC_MODULE)"
1045 code (in function swcr_process) so that it compiled if these where
1046 not defined.
1047
1048 2005-11-11 01:44 davidm
1049
1050 * Makefile, README:
1051
1052 cleanups for a general release of OCF
1053
1054 2005-11-11 01:18 davidm
1055
1056 * patches/ssl.patch:
1057
1058 Updated the ssl patch to the latest and greatest changed we have
1059 made.
1060
1061 2005-11-10 21:41 davidm
1062
1063 * Makefile, ixp4xx/ixp4xx.c:
1064
1065 Better debug for bad input.
1066
1067 Have make clean do more cleaning and less talking.
1068
1069 2005-11-03 20:53 davidm
1070
1071 * cryptosoft.c:
1072
1073 clean up some compilation errors with various options on/off
1074
1075 2005-10-25 00:25 davidm
1076
1077 * Makefile:
1078
1079 Add a patch target that generates full kernel patches to add OCF
1080 into either a 2.4 or 2.6 kernel as a single patch.
1081
1082 2005-10-25 00:24 davidm
1083
1084 * Kconfig:
1085
1086 Make sure all OCF settings depend on OCF_OCF
1087
1088 2005-09-23 02:45 davidm
1089
1090 * README, README.sglinux:
1091
1092 new crypto-tools archive to keep the tools up to date
1093
1094 2005-09-23 02:08 davidm
1095
1096 * Makefile, README, README.sglinux:
1097
1098 updates for doing OCF releases
1099
1100 2005-09-23 01:59 davidm
1101
1102 * patches/: ssh.patch, ssl.patch:
1103
1104 updated the patches for the latest in fixes etc to ssh/ssl for a
1105 new OCF release before the openswan 2.4.1 merge.
1106
1107 2005-09-21 00:57 davidm
1108
1109 * Makefile, cryptosoft.c, hifn/hifn7751.c, ixp4xx/ixp4xx.c:
1110
1111 Fixes for building cleanly under 2.6
1112
1113 2005-09-13 23:11 davidm
1114
1115 * ocf-bench.c:
1116
1117 Add an exit function for cleaner 2.6 support. Patch from Ronen
1118 Shitrit
1119
1120 2005-07-30 00:23 davidm
1121
1122 * cryptosoft.c:
1123
1124 Add proper hooks for MD5/SHA and their HMAC countrparts processing
1125 from cryptodev.
1126
1127 2005-07-29 01:50 davidm
1128
1129 * cryptodev.c:
1130
1131 cryptodev did not support MD5 and SHA1, only the HMAC versions
1132
1133 2005-07-29 01:05 davidm
1134
1135 * ixp4xx/ixp4xx.c:
1136
1137 register for MD5, return 16 bytes for MD5 and 12 for MD5_HMAC,
1138 likewise for SHA1 (only 20 is not HMAC).
1139
1140 2005-07-28 21:52 davidm
1141
1142 * ixp4xx/ixp4xx.c:
1143
1144 First pass fix of userland MD5 processing. We now do as well as
1145 the safenet does, I think.
1146
1147 2005-07-27 06:23 davidm
1148
1149 * cryptodev.c:
1150
1151 udelay doesn't give up the current thread, thus the kernel will get
1152 locked if a process is killed but the hardware never completes the
1153 crypto request.
1154
1155 2005-07-22 02:07 davidm
1156
1157 * crypto.c, cryptodev.h:
1158
1159 Implement queuing limits for input/output and OCF requests.
1160
1161 Implement 2.6 style work queues instead of the 2.4 task_queues.
1162
1163 2005-07-21 20:42 davidm
1164
1165 * cryptodev.h, ocf-bench.c, ixp4xx/ixp4xx.c:
1166
1167 Fix OCF to use work queues, add 2.4 version of work queues to
1168 cryptodev.h for use in ipsec and elsewhere.
1169
1170 Problem and initial patch provided by David Vrabel
1171 . Cleaned up and 2.4 support added by me.
1172
1173 2005-07-21 19:08 davidm
1174
1175 * random.c:
1176
1177 Fix spinlock initialisation, problem reported by Andy @
1178 windsorcarclub.co.uk.
1179
1180 2005-07-20 20:24 davidm
1181
1182 * cryptodev.c:
1183
1184 fix a silly spelling mistake
1185
1186 2005-07-08 00:56 gerg
1187
1188 * Makefile:
1189
1190 Only build ocf-bench when CONFIG_OCF_BENCH is acrually enabled.
1191
1192 2005-06-27 20:29 davidm
1193
1194 * ixp4xx/ixp4xx.c:
1195
1196 Ok, we need a simple implentation here or we go too slow for UDP
1197 tests. For now, if the Q is full just ditch the packet, someone
1198 has to do it.
1199
1200 2005-06-25 01:13 davidm
1201
1202 * safe/safe.c:
1203
1204 bytes swapping etc all seems wrong for safenet on BE, these fixes
1205 see both MD5 and SHA1 working with OpenSwan.
1206
1207 2005-06-22 23:10 davidm
1208
1209 * random.c:
1210
1211 clean up some ifdef code a little
1212
1213 2005-06-22 21:28 davidm
1214
1215 * ixp4xx/ixp4xx.c:
1216
1217 Make sure we do not call blocking functions from ISR callable
1218 routines. In this case we were calling ixCryptoAccCtxUnregister.
1219
1220 Run all the random numbers through SHA1 process to ensure more
1221 uniform distribution of bits (NOTE: it is not more random in any
1222 way)
1223
1224 2005-06-21 00:11 davidm
1225
1226 * ixp4xx/ixp4xx.c:
1227
1228 do not process requests from the register callback, gets the AES
1229 code all messed up.
1230
1231 Align caches on HW cache boundaries ot improve speed.
1232
1233 More tracking or potential errors.
1234
1235 2005-06-15 01:55 davidm
1236
1237 * ixp4xx/ixp4xx.c:
1238
1239 technically ixCryptoAccCtxRegister cannot be called from IRQ
1240 context, so run it from the immediate BH.
1241
1242 2005-06-14 23:13 davidm
1243
1244 * ocf-bench.c:
1245
1246 Fix some compile warnings/errors
1247
1248 2005-06-14 20:52 davidm
1249
1250 * Config.in, Kconfig, Makefile, ocf-bench.c:
1251
1252 Add in kernel benchmark driver
1253
1254 Support for building under CSR 1.4 and 2.0 is now complete
1255
1256 2005-06-14 20:51 davidm
1257
1258 * crypto.c:
1259
1260 Do not use immediate processing for SW drivers otherwise we hold a
1261 lock for too long. Instead force BATCH processing.
1262
1263 Problem found by David Vrabel
1264
1265 2005-06-14 20:46 davidm
1266
1267 * cryptodev.c:
1268
1269 Fix up AES minimum key size
1270
1271 Make some more variables static
1272
1273 2005-06-14 20:36 davidm
1274
1275 * ixp4xx/ixp4xx.c:
1276
1277 Use kernel caches as they are more efficient and faster to obtain.
1278
1279 Fix some spelling
1280
1281 Tune PKE to only use the space required. Turn off go fast options
1282 to reduce any speed-related cracking.
1283
1284 Only zero data that needs to be zeroed (save some cycles)
1285
1286 2005-06-02 21:42 davidm
1287
1288 * ixp4xx/ixp4xx.c:
1289
1290 Added optimisations ideas from Intel the improve the PKE
1291 performance for 512 and 1024 bits operations.
1292
1293 2005-06-01 02:13 davidm
1294
1295 * ixp4xx/ixp4xx.c:
1296
1297 Looks like I found the ixp bug. Using OSAL buffer routines on
1298 non-OSAL buffers is a very very bad thing to do. Must double check
1299 all the API's I am using (ie., PKE) just to be sure.
1300
1301 2005-05-31 21:38 davidm
1302
1303 * Config.in, Kconfig:
1304
1305 Updated/Added the menu wrapper for the config options
1306
1307 2005-05-31 21:18 gerg
1308
1309 * Config.in:
1310
1311 Reworked the config.in so that each sub-module has its own
1312 Config.in. That way it is easier to make release trees with some
1313 modules left in.
1314
1315 2005-05-30 19:46 davidm
1316
1317 * ixp4xx/ixp4xx.c:
1318
1319 Not all the Pke code was appropriately ifdef'd
1320
1321 2005-05-28 01:49 davidm
1322
1323 * cryptosoft.c:
1324
1325 We were not injecting data in the right places nor calling the
1326 cipher code the best was under all situations. We are now good
1327 enought to do ESP/AH processing with 3DES and SHA1-HMAC.
1328
1329 2005-05-28 01:42 davidm
1330
1331 * hifn/hifn7751.c:
1332
1333 Non atomic kmallocs at IRQ time are bad and cause lockups
1334
1335 2005-05-21 08:31 davidm
1336
1337 * README, README.sglinux:
1338
1339 Some small updates to email and patches that no longer exist
1340
1341 2005-05-21 08:25 davidm
1342
1343 * cryptosoft.c:
1344
1345 better error message on failure, a lot of the error check we do
1346 sems to break apps like cryptotest, not sure who is right yet.
1347
1348 2005-05-21 00:55 davidm
1349
1350 * criov.c, crypto.c, cryptodev.c, cryptodev.h, cryptosoft.c,
1351 random.c, rndtest.c, uio.h, hifn/hifn7751.c, ixp4xx/ixp4xx.c,
1352 safe/safe.c, safe/safevar.h:
1353
1354 Convert to CyberGuard email addresses for OCF files.
1355
1356 2005-05-21 00:28 davidm
1357
1358 * crypto.c, cryptodev.c, cryptosoft.c, hifn/hifn7751.c,
1359 ixp4xx/ixp4xx.c, safe/safe.c:
1360
1361 final round of naming changes, hifn driver also confirmed working
1362 in SG710
1363
1364 2005-05-20 23:50 davidm
1365
1366 * cryptosoft.c, random.c, uio.h, ixp4xx/ixp4xx.c:
1367
1368 more copyright/author updates etc for Intel
1369
1370 2005-05-20 23:48 davidm
1371
1372 * criov.c, crypto.c, cryptodev.c, cryptodev.h, cryptosoft.c,
1373 random.c, rndtest.c, uio.h, ixp4xx/ixp4xx.c, safe/safe.c,
1374 safe/safevar.h:
1375
1376 updated copyrights to reflect Intels interest/investment in the
1377 project
1378
1379 2005-05-20 20:31 davidm
1380
1381 * hifn/: hifn7751.c, hifn7751var.h:
1382
1383 Most of a working SKB implementation now
1384
1385 2005-05-20 20:30 davidm
1386
1387 * safe/: md5.c, md5.h, safe.c, sha1.c, sha1.h:
1388
1389 Full AH/ESP processing now working, just added BSD sha/md5
1390 routines as needed for now.
1391
1392 2005-05-20 20:30 davidm
1393
1394 * random.c:
1395
1396 use the new BSD rndtest code rather than the old GPL fips code
1397
1398 2005-05-20 20:28 davidm
1399
1400 * Makefile, fips.c, fips.h, rndtest.c, rndtest.h:
1401
1402 Get rid of the GPL version of the fips test and use a BSD licensed
1403 version so no one can get upset and cry derivative :-)
1404
1405 2005-05-20 08:19 davidm
1406
1407 * ixp4xx/ixp4xx.c:
1408
1409 High throughput improvements. Can now handle multiple simultaneous
1410 requests.
1411
1412 2005-05-20 00:55 davidm
1413
1414 * ixp4xx/ixp4xx.c:
1415
1416 works with openswan/OCF for ipsec receive.
1417
1418 fixes ESP/AUTH processing (esp. HMAC)
1419
1420 This driver needs a rework for kernel IPSEC as it's current packet
1421 at a time processing is not up to the task.
1422
1423 2005-05-20 00:53 davidm
1424
1425 * cryptosoft.c:
1426
1427 working ESP/AUTH code now impleneted. Can do RX processing for
1428 openswan KLIPS.
1429
1430 Fixes numerous problems in skb processing
1431
1432 Fixes broken HMAC code and IV processing
1433
1434 2005-05-14 01:44 davidm
1435
1436 * cryptodev.c, ixp4xx/ixp4xx.c:
1437
1438 Cleanup IXP key processing to guarantee only a single outstanding
1439 request rather than relying on the intel driver to get it right.
1440 Stops us losing requests.
1441
1442 Tighten up the cryptodev response to "no answer" so that we don't
1443 take all the CPU. Only happens if there is a driver bug.
1444
1445 2005-05-14 00:07 davidm
1446
1447 * Config.in, Kconfig, Makefile, crypto.c, fips.c, fips.h, random.c,
1448 ixp4xx/ixp4xx.c, tools/bench-ocf:
1449
1450 PKE and RNG support running on the ixp driver, added PKE bench
1451 marking to script. Still some multi-thread problems in the PKE
1452 code.
1453
1454 Added FIP RNG checking option to config and code.
1455
1456 2005-05-10 19:18 davidm
1457
1458 * Makefile, ixp4xx/ixp4xx.c:
1459
1460 Get the OCF stuff building for the 465 and CSR-2.0
1461
1462 2005-04-27 19:18 davidm
1463
1464 * cryptodev.h, random.c, hifn/hifn7751.c, safe/safe.c:
1465 cleanup the random number interface some more
1466
1467 2005-04-27 00:57 davidm
1468
1469 * cryptodev.h, cryptosoft.c, random.c, hifn/hifn7751.c,
1470 ixp4xx/ixp4xx.c, patches/linux-2.4.29-add_true_randomness.patch,
1471 patches/linux-2.6.11-add_true_randomness.patch, safe/safe.c,
1472 safe/safevar.h:
1473
1474 Switch to a more "user" like random number handling. Drivers no
1475 longer poll for RNG data, we pull it as needed to fill
1476 /dev/random's entropy.
1477
1478 Implement sk_buff handling within the OCF framework.
1479
1480 fixup IV handling in cryptosoft.
1481
1482 2005-04-27 00:41 davidm
1483
1484 * crypto.c:
1485
1486 Fix a race condition with the starting of kernel threads. The
1487 threads were running before the pid assignment in the parent.
1488
1489 2005-03-24 23:57 davidm
1490
1491 * Makefile:
1492 include crypto-tools.patch in the release file
1493
1494
1495
1496 2005-03-24 20:14 davidm
1497
1498 * safe/safe.c:
1499 Fixup compile time warnings due to some left over BSDisms
1500
1501
1502 2005-03-24 00:53 davidm
1503
1504 * Makefile, README, README.sglinux, patches/linux-2.4.29-ocf.patch,
1505 patches/linux-2.6.11-ocf.patch, patches/ocf-linux-2.4.29.patch,
1506 patches/ocf-linux-2.6.11.patch:
1507 added cleaner patch names and a tarball target to aid releases
1508
1509
1510 2005-03-24 00:28 davidm
1511
1512 * patches/crypto-tools.patch,
1513 patches/linux-2.4.29-add_true_randomness.patch,
1514 patches/linux-2.6.11-add_true_randomness.patch,
1515 patches/ocf-linux-2.4.29.patch, patches/ocf-linux-2.6.11.patch,
1516 patches/ssh.patch, patches/ssl.patch, tools/bench-ocf:
1517 move all the release file patches into CVS for simplicity
1518
1519
1520 2005-03-23 20:37 davidm
1521
1522 * safe/safe.c:
1523 remove excessive debug from RNG routines so that you can turn on
1524 debug and live through it
1525
1526
1527
1528 2005-03-23 02:23 davidm
1529
1530 * safe/safe.c:
1531 fix memory corruption for mod_exp and the safenet, we were copying
1532 back more than the space available.
1533
1534
1535
1536 2005-03-22 21:45 davidm
1537
1538 * crypto.c, cryptodev.c, cryptodev.h, safe/safe.c:
1539 fixup a major sync issues with key processing (callback called
1540 before sleeping). Improve its performance while we are there with
1541 a CBIMM (callback immediate) option.
1542
1543
1544
1545 2005-03-19 00:33 davidm
1546
1547 * random.c:
1548 A new randomness function for both 2.4 and 2.6 that replaces out
1549 previous old solution for the hifn driver with more generic code
1550 that works on both kernels.
1551
1552 add_true_randomness(u_int32_t *buf, int nwords);
1553
1554
1555
1556 2005-03-18 21:01 davidm
1557
1558 * Makefile, cryptodev.h, random.c, hifn/hifn7751.c, safe/safe.c:
1559 RNG support in both the safenet and the hifn plus the required
1560 kernel support.
1561
1562 "hd /dev/random" runs much much faster now :-)
1563
1564
1565
1566 2005-03-17 23:29 toby
1567
1568 * cryptodev.c: Make sure the CIOCASYMFEAT ioctl on
1569 /dev/crypto copies out the capable features.
1570
1571 2005-03-17 01:19 davidm
1572
1573 * safe/: safe.c, safereg.h, safevar.h:
1574 hardware PK acceleration on the safenet (CRK_MOD_EXP only)
1575
1576
1577 2005-03-16 04:28 davidm
1578
1579 * criov.c, crypto.c, cryptodev.c, safe/safe.c:
1580 fixup the FBSD id stuff to compile :-)
1581
1582 2005-03-16 04:02 davidm
1583
1584 * README, README.sglinux, TODO:
1585 Updated with versions, removed tabs, new kernel versions, web
1586 site etc
1587
1588 2005-03-16 03:45 davidm
1589
1590 * criov.c, crypto.c, cryptodev.c, cryptodev.h:
1591 more license formatting and version id's to help keep in touch with
1592 FreeBSD
1593
1594 2005-03-16 03:16 davidm
1595
1596 * safe/safe.c:
1597 up to date with FreeBSD versioning, no changes to merge
1598
1599 2005-03-16 03:15 davidm
1600
1601 * safe/safe.c:
1602 include FreeBSD version info so I can track changes better
1603
1604
1605 2005-03-16 03:11 davidm
1606
1607 * hifn/: hifn7751.c, hifn7751reg.h, hifn7751var.h:
1608 incorporate latest updates from FreeBSD:
1609
1610 | Update support for 795x parts: | o rework pll setup code to
1611 follow h/w specification | o add hint.hifn.X.pllconfig to specify
1612 reference clock setup | requirements; default is pci66 which
1613 means the clock is | derived from the PCI bus clock and the card
1614 resides in a | 66MHz slot | | Tested on 7955 and 7956 cards;
1615 support for 7954 cards not enabled | since we have no cards to test
1616 against.
1617
1618
1619
1620 2005-03-16 02:50 davidm
1621
1622 * Config.in, Kconfig, README, crypto.c, cryptodev.c, cryptodev.h,
1623 hifn/hifn7751.c, hifn/hifn7751var.h, safe/safe.c, safe/safevar.h:
1624 cleanups to build and run on 2.6.11 and 2.4.29 for a public release
1625
1626 included configs locally to reduce patch to kernel and required
1627 user effort.
1628
1629 pci_dma_sync_single fix from Michele Baldessari
1630 , with modifications by me to work on all
1631 kernels.
1632
1633
1634
1635 2005-03-11 23:30 davidm
1636
1637 * Makefile:
1638 fix up the compiling again, I had outstanding commits for this one
1639 :-)
1640
1641 2005-03-11 21:56 danield
1642
1643 * Makefile: Get ocf modules building again.
1644
1645 2004-12-25 07:12 davidm
1646
1647 * TODO:
1648 safenet is all good now AKAIK
1649
1650 2004-12-25 07:02 davidm
1651
1652 * Makefile, crypto.c, cryptodev.c, cryptosoft.c, uio.h:
1653 2.6 port of the OCF framework, tested on Xscale and compiled for
1654 x86
1655
1656
1657
1658 2004-12-10 02:41 davidm
1659
1660 * hifn/hifn7751.c:
1661 Fix compilation as l_flags was not defined for LOCK in pci_remove
1662
1663
1664 2004-12-06 19:15 davidm
1665
1666 * hifn/hifn7751.c, safe/safe.c:
1667 * Put locking into hifn_remove to ensure interrupts are not running
1668 while we remove the driver
1669
1670 Use del_timer_sync (need to ensure timer is not running on
1671 another CPU
1672 * when we delete it).
1673
1674 Improvements suggested by Evgeniy Polyakov <johnpol@2ka.mipt.ru>
1675
1676
1677 2004-12-02 09:16 davidm
1678
1679 * README, README.sglinux:
1680 * Fix up all tabs to be spaces * explain how to update the ocf
1681 support in the patch to the current version.
1682
1683
1684
1685 2004-12-02 09:11 davidm
1686
1687 * README:
1688 * fix some spelling/grammar * add more info on the ssl.patch file
1689 and what it contains
1690
1691 2004-12-02 09:08 davidm
1692
1693 * README.sglinux:
1694 give some more instructions on the shar archive and extracting the
1695 release.
1696
1697 2004-12-02 09:03 davidm
1698
1699 * README.sglinux:
1700 how to include crypto-tools into 3.1.6
1701
1702
1703
1704 2004-12-02 08:48 davidm
1705
1706 * README.sglinux:
1707 updated for a new crypto patch for openssl-0.9.7e
1708
1709 2004-12-02 08:04 davidm
1710
1711 * README, README.sglinux:
1712 Clean up the README's to use the new SG Linux alpha and a specific
1713 2.4.28 patch.
1714
1715
1716
1717 2004-12-01 23:11 davidm
1718
1719 * Makefile, criov.c, crypto.c, cryptodev.c, cryptosoft.c,
1720 hifn/Makefile, hifn/hifn7751.c, ixp4xx/Makefile, ixp4xx/ixp4xx.c,
1721 safe/Makefile, safe/safe.c:
1722 Re-worked the Makefiles so that including the code into a standard
1723 non-SG kernel is really easy.
1724
1725 Fixed a non-initialised bug that was introduce into cryptosoft.c
1726 with the additional error checking.
1727
1728
1729
1730 2004-12-01 01:54 davidm
1731
1732 * TODO:
1733 updates based on things that have been fixed.
1734
1735 2004-12-01 01:53 davidm
1736
1737 * safe/: safe.c, safevar.h:
1738 Cleanup some old hacks based on the much cleaner port of the hifn
1739 driver.
1740
1741 2004-12-01 01:53 davidm
1742
1743 * cryptosoft.c:
1744 fix some crashes due to bad buffer sizes etc.
1745
1746 2004-11-27 09:16 davidm
1747
1748 * crypto.c, cryptodev.c, ixp4xx/ixp4xx.c:
1749 Fixed some nasty problems that were stopping reliable behaviour:
1750
1751 1) we were not initialising some of our lists/wait queues, which
1752 meant they appears to have things in them when in fact they did
1753 not. This actually didn't seems to cause any problems but is
1754 extremely bogus.
1755
1756 2) While a process was waiting for the crypto engine, if
1757 itreceived a signal we would lose sync with the engine and end
1758 up allowing out-of-band actions that were invalid (ie.,
1759 unregistering a context that was still active in a crypto
1760 operation).
1761
1762 Of source the CryptACC should probably deal with this as well
1763 ;-)
1764
1765
1766
1767 2004-11-26 01:33 davidm
1768
1769 * TODO:
1770 We should also hook in the random number generators to linux
1771
1772
1773 2004-11-26 01:33 davidm
1774
1775 * hifn/hifn7751.c:
1776 Changed all the accesses to DMS descriptors to not set the valid
1777 bit until after everything else was set. This got the driver
1778 running smoothly, along with a fixup to the pci_map_uio which
1779 wasn't settings lengths correctly.
1780
1781
1782
1783 2004-11-25 21:15 davidm
1784
1785 * ixp4xx/ixp4xx.c:
1786 document why using a new context for each packet is actually a
1787 better idea than allocating one (actually two) per session.
1788
1789
1790
1791 2004-11-25 08:48 davidm
1792
1793 * README, README.sglinux:
1794 updates from email with Intel to fix a few little things
1795
1796
1797 2004-11-25 00:02 davidm
1798
1799 * README, README.sglinux, safe/safe.c:
1800 Some cleanups of doc and so on for Intel/General use
1801
1802 2004-11-23 07:58 davidm
1803
1804 * TODO:
1805 safenet is working on big endian machines now
1806
1807
1808
1809 2004-11-23 07:56 davidm
1810
1811 * ixp4xx/: Makefile, ixp4xx.c:
1812 Reference the Intel library, cleanup the IV sizes and turn on AES.
1813 Fix Makefile for new config options.
1814
1815
1816
1817 2004-11-23 07:52 davidm
1818
1819 * Makefile:
1820 Complete the changes of ixp to ixp4xx (IXP4xx) as appropriate
1821
1822
1823 2004-11-23 00:33 davidm
1824
1825 * Makefile, README, ixp4xx/Makefile:
1826 Changes all references to ixp to be ixp4xx at Intels request
1827
1828
1829 2004-11-20 01:07 davidm
1830
1831 * safe/: safe.c, safereg.h:
1832 fully working safenet on BE machines
1833
1834 2004-11-19 01:03 davidm
1835
1836 * Makefile, hifn/Makefile, hifn/hifn7751.c, hifn/hifn7751reg.h,
1837 hifn/hifn7751var.h:
1838 hifn driver code complete and compiling, needs a test now ;-)
1839
1840
1841 2004-11-18 21:45 davidm
1842
1843 * hifn/: Makefile, hifn7751.c, hifn7751var.h:
1844 Makefile for building the hifn driver
1845
1846 2004-11-18 21:44 davidm
1847
1848 * hifn/: hifn7751.c, hifn7751reg.h, hifn7751var.h:
1849 Checkin the orginal freebsd source for the hifn driver as a
1850 reference in the future.
1851
1852
1853
1854 2004-11-18 10:05 davidm
1855
1856 * Makefile, README, TODO, criov.c, crypto.c, cryptodev.c,
1857 cryptodev.h, cryptosoft.c, uio.h, ixp4xx/Makefile, ixp4xx/ixp4xx.c,
1858 safe/Makefile, safe/safe.c, safe/safevar.h:
1859 Clean up license and copyright info to be more acceptable (and
1860 present) Clean up debug and trace Fixup memory freeing etc on
1861 safenet Fix compiler warnings (some were bugs)
1862
1863
1864
1865 2004-11-17 02:23 davidm
1866
1867 * safe/safe.c:
1868 working on Xscale (big endian) now but packet data is getting
1869 stuffed up due to endian problems (at least now we are talking to
1870 the chip correctly for BE). Good enough to test packet throughput,
1871 no good for testing scp.
1872
1873
1874
1875 2004-11-17 02:21 davidm
1876
1877 * criov.c:
1878 Make sure public symbols are exported by including the correct
1879 header files
1880
1881 2004-11-17 02:15 davidm
1882
1883 * crypto.c:
1884 even better cleanup of kernel threads on exit
1885
1886 2004-11-17 02:15 davidm
1887
1888 * cryptosoft.c:
1889 return some trace to debug so it doesn't appear
1890
1891 2004-11-17 02:14 davidm
1892
1893 * Makefile:
1894 Make sure all drivers get built, not just IXP
1895
1896 2004-11-16 21:31 davidm
1897
1898 * crypto.c:
1899 Fix problem with reboots and driver not unloading cleanly, we were
1900 not handling signals correctly in the kernel threads,
1901
1902 2004-11-10 10:46 davidm
1903
1904 * ixp4xx/ixp4xx.c:
1905 fix serious context leak, itturns out the context is still
1906 considered busy while it is calling the perform callback, so we
1907 cleanup on closing the session and on allocating the next context
1908 to work around this.
1909
1910
1911
1912 2004-11-10 05:26 davidm
1913
1914 * crypto.c, cryptodev.c, ixp4xx/ixp4xx.c, safe/safe.c:
1915 cleaned out some debug, found MAX tunnels bug, traced it various
1916 other cleanups.
1917
1918
1919
1920 2004-11-10 04:02 davidm
1921
1922 * cryptodev.c, cryptosoft.c, ixp4xx/ixp4xx.c, safe/Makefile,
1923 safe/safe.c:
1924 Fix up kmalloc usage to always zero resulting buffer everywhere
1925 (stops crashes in ixp)
1926
1927 Add some function debug to ixp so you can see it working
1928
1929 Fix safe driver to build and install in a real tree.
1930
1931
1932
1933 2004-11-10 02:27 davidm
1934
1935 * Makefile, criov.c, crypto.c, cryptosoft.c, ixp4xx/Makefile,
1936 ixp4xx/ixp4xx.c:
1937 Compiling OCF modules for the IXP crypto, needs testing now
1938
1939
1940 2004-11-09 19:16 davidm
1941
1942 * criov.c, crypto.c, cryptodev.c, cryptodev.h, cryptosoft.c,
1943 cryptosoft.h, uio.h, safe/safe.c, safe/safevar.h:
1944 The linux port of OCF with working safenet and software modules.
1945 Still some bugs with mutiple crypto threads using the safenet
1946 driver.
1947
1948
1949
1950 2004-11-09 18:49 davidm
1951
1952 * Makefile, criov.c, crypto.c, cryptodev.c, cryptodev.h,
1953 cryptosoft.c, cryptosoft.h, uio.h, safe/Makefile, safe/safe.c,
1954 safe/safereg.h, safe/safevar.h:
1955 Check in the orginal free-bsd sources for the OCF support. This
1956 allows us to diff against it later to see if we botched anything
1957 major league.
1958
1959
1960
This page took 0.139591 seconds and 5 git commands to generate.