Yavor Goulishev | bda86ae | 2011-05-23 11:56:53 -0700 | [diff] [blame] | 1 | commit a04cbb0095a78aeed3f65aaf06c46069d0ac184a |
| 2 | Author: Daniel Drake <dan@reactivated.net> |
| 3 | Date: Tue May 4 19:01:59 2010 -0300 |
| 4 | |
| 5 | v1.0.8 release |
| 6 | |
| 7 | commit 53b47299531974fa8901b74a163f3c8ebae4eec9 |
| 8 | Author: Daniel Drake <dan@reactivated.net> |
| 9 | Date: Thu Apr 22 18:12:51 2010 -0300 |
| 10 | |
| 11 | Linux: don't set SHORT_NOT_OK on bulk out URBs (#20 |
| 12 | |
| 13 | Setting this flag is illegal, and the behaviour we're looking for |
| 14 | is already in place for host-to-device transfers without this flag. |
| 15 | |
| 16 | commit 1519828b7ee1fce46d4c51fc097d52e01b8e0bb4 |
| 17 | Author: Daniel Drake <dan@reactivated.net> |
| 18 | Date: Tue Apr 20 20:14:09 2010 -0300 |
| 19 | |
| 20 | Linux: Handle failure to read active configuration during enumeration |
| 21 | |
| 22 | commit 45168627cc15aee3875192f34286110dbbd27095 |
| 23 | Author: Nathan Hjelm <hjelmn@me.com> |
| 24 | Date: Tue Apr 20 19:43:06 2010 -0300 |
| 25 | |
| 26 | Darwin: don't reuse cached descriptors during enumeration |
| 27 | |
| 28 | Descriptor reuse was causing scans to return invalid information when |
| 29 | the device at a location has changed. |
| 30 | |
| 31 | commit 3af329db6dcbfb96d24867c757f9e125e5b7b0e5 |
| 32 | Author: Daniel Drake <dan@reactivated.net> |
| 33 | Date: Mon Apr 19 19:36:55 2010 -0300 |
| 34 | |
| 35 | v1.0.7 release |
| 36 | |
| 37 | commit ceb8cacd6d2a4189de0db2ee46d45217511c69be |
| 38 | Author: Daniel Drake <dan@reactivated.net> |
| 39 | Date: Mon Apr 19 19:31:12 2010 -0300 |
| 40 | |
| 41 | Document that fd set contents are an internal implementation detail (#36) |
| 42 | |
| 43 | commit 9bea500b5747bdeba7c8251d45608558e71a1db5 |
| 44 | Author: Martin Koegler <mkoegler@auto.tuwien.ac.at> |
| 45 | Date: Mon Apr 19 19:22:46 2010 -0300 |
| 46 | |
| 47 | Linux: correct config descriptor endianness fix |
| 48 | |
| 49 | The seek_to_next_config codepath deals with both sysfs and usbfs; |
| 50 | make sure we only convert values in the usbfs path. |
| 51 | |
| 52 | commit 02df59a309e813c50b8230de99e69fb4e1814279 |
| 53 | Author: Martin Koegler <mkoegler@auto.tuwien.ac.at> |
| 54 | Date: Mon Apr 5 17:11:32 2010 +0200 |
| 55 | |
| 56 | Linux: Fix endianness handling of usbfs config descriptors (#27) |
| 57 | |
| 58 | driver/usb/core/devio.c function usbdev_read translate the follwing files |
| 59 | to CPU endianess: |
| 60 | |
| 61 | le16_to_cpus(&temp_desc.bcdUSB); |
| 62 | le16_to_cpus(&temp_desc.idVendor); |
| 63 | le16_to_cpus(&temp_desc.idProduct); |
| 64 | le16_to_cpus(&temp_desc.bcdDevice); |
| 65 | |
| 66 | All other data is passed without any change. |
| 67 | |
| 68 | libusb/os/linux_usbfs.c calls usbi_parse_descriptor with host_endian=1 |
| 69 | for config descriptors. According to the kernel code, they must be |
| 70 | processed with host_endian=0, as they are not translated by the kernel. |
| 71 | |
| 72 | Signed-off-by: Martin Koegler <mkoegler@auto.tuwien.ac.at> |
| 73 | |
| 74 | commit ec303b01a5d4e51c000a283853af65059fa62285 |
| 75 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 76 | Date: Tue Mar 23 16:22:08 2010 -0600 |
| 77 | |
| 78 | Darwin: fix enumeration of devices with non-consecutive addresses (#23) |
| 79 | |
| 80 | commit cd809e2f7cee3874b7ae16b2c482a8b63a90e4a5 |
| 81 | Author: Bastien Nocera <hadess@hadess.net> |
| 82 | Date: Mon Mar 8 10:50:51 2010 +0000 |
| 83 | |
| 84 | Add more interface classes |
| 85 | |
| 86 | As used in bluez. |
| 87 | |
| 88 | commit 1ce4aa67d849f5cad8a21072dc1c7b42158ce817 |
| 89 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 90 | Date: Thu Mar 18 11:32:32 2010 -0600 |
| 91 | |
| 92 | Darwin: Cleanup async callback code, catch request timeouts |
| 93 | |
| 94 | commit 68af9f8d731f700267335941a8214d34ab518cc8 |
| 95 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 96 | Date: Thu Mar 18 11:31:46 2010 -0600 |
| 97 | |
| 98 | Darwin: fix endianness of control setup packet |
| 99 | |
| 100 | IOUSBLib expects the control request to be in host byte order. |
| 101 | Swap the request into host byte order. |
| 102 | |
| 103 | commit cfce4d127184f4e4f334976151a0f80594bb5e22 |
| 104 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 105 | Date: Mon Feb 15 14:10:08 2010 -0600 |
| 106 | |
| 107 | Darwin: fix memory leak in process_device |
| 108 | |
| 109 | Credit to Mike Colagrosso for finding this bug. |
| 110 | |
| 111 | commit 161893cfbefefe315f657677705abe090fc526f2 |
| 112 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 113 | Date: Mon Feb 15 14:09:19 2010 -0600 |
| 114 | |
| 115 | Darwin: use logging functions |
| 116 | |
| 117 | Use usbi_warn, usbi_err, and usbi_info instead of _usbi_log. |
| 118 | |
| 119 | commit 2a72f38548208044dc3aa62681419d006c35732d |
| 120 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 121 | Date: Mon Feb 15 14:07:44 2010 -0600 |
| 122 | |
| 123 | Darwin: support multiple calls to libusb_init |
| 124 | |
| 125 | Credit to Orin Eman for finding this bug. |
| 126 | |
| 127 | commit a4186794d87124503db2f5f51f51ce90bb95daa7 |
| 128 | Author: Daniel Drake <dan@reactivated.net> |
| 129 | Date: Sun Nov 22 17:20:53 2009 +0000 |
| 130 | |
| 131 | v1.0.6 release |
| 132 | |
| 133 | commit 8392ff22136fccaf1e15d186157609b8dd127bc5 |
| 134 | Author: Ludovic Rousseau <rousseau@debian.org> |
| 135 | Date: Sun Oct 25 10:06:41 2009 +0100 |
| 136 | |
| 137 | lsusb example: make print_devs() static |
| 138 | |
| 139 | lsusb.c:26: warning: no previous prototype for ‘print_devs’ |
| 140 | |
| 141 | commit 9cc6bfaa15239bb6db1c1570b9beb6df2f848951 |
| 142 | Author: Ludovic Rousseau <rousseau@debian.org> |
| 143 | Date: Sun Oct 25 10:05:10 2009 +0100 |
| 144 | |
| 145 | Darwin: fix warning in darwin_error_str() |
| 146 | |
| 147 | os/darwin_usb.c:63: warning: return discards qualifiers from pointer |
| 148 | target type |
| 149 | |
| 150 | commit 4c706d2fb6b2c43b10d72ac5dff51cac4d939f1a |
| 151 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 152 | Date: Sat Nov 21 17:06:43 2009 +0000 |
| 153 | |
| 154 | Darwin: allow devices to be opened multiple times |
| 155 | |
| 156 | Allows libusb applications to access multiple interfaces of the same |
| 157 | device in the same application. |
| 158 | |
| 159 | Also fixes a set alt interface bug. |
| 160 | |
| 161 | commit 0232fc559cdacb9561f982dd6d28feb4435b3e4e |
| 162 | Author: Daniel Drake <dan@reactivated.net> |
| 163 | Date: Sat Nov 21 17:01:32 2009 +0000 |
| 164 | |
| 165 | Increase libusb_handle_events() timeout to 60 seconds |
| 166 | |
| 167 | The internal timing seems to be working, this will be a better test of |
| 168 | it before we make this timeout unlimited. |
| 169 | |
| 170 | commit 0bd7ef5d8697973a026c36c15d6276177b4ec4ea |
| 171 | Author: Daniel Drake <dan@reactivated.net> |
| 172 | Date: Sat Nov 21 16:57:25 2009 +0000 |
| 173 | |
| 174 | Refine timerfd header check (#18) |
| 175 | |
| 176 | Require glibc-2.9 for the working timerfd support. |
| 177 | |
| 178 | commit 90d8fcab9018b8e6887a7e0592d1e5f692117234 |
| 179 | Author: Daniel Drake <dsd@gentoo.org> |
| 180 | Date: Sun Nov 15 12:17:13 2009 +0000 |
| 181 | |
| 182 | v1.0.5 release |
| 183 | |
| 184 | commit 11d591058e3f105b0e90c23bbf58b18de691e690 |
| 185 | Author: Daniel Drake <dsd@gentoo.org> |
| 186 | Date: Sat Nov 7 10:43:59 2009 +0000 |
| 187 | |
| 188 | Update documentation about early completion caveats |
| 189 | |
| 190 | commit f796c9528a71aa55326b6f0c9c7c5ec073d2bf92 |
| 191 | Author: Ludovic Rousseau <rousseau@debian.org> |
| 192 | Date: Sun Oct 25 09:59:51 2009 +0100 |
| 193 | |
| 194 | Add libusb_get_max_iso_packet_size() to libusb.h |
| 195 | |
| 196 | core.c:777: warning: no previous prototype for 'libusb_get_max_iso_packet_size' |
| 197 | |
| 198 | commit 4783008b7e711de9cb31631e60dda995f44068de |
| 199 | Author: Daniel Drake <dsd@gentoo.org> |
| 200 | Date: Wed Oct 28 20:33:49 2009 +0545 |
| 201 | |
| 202 | Use timerfd for timeout handling |
| 203 | |
| 204 | Use a new file descriptor from the timerfd system calls to handle |
| 205 | timeouts. On supported systems, this means that there is less hassle |
| 206 | figuring out when the poll() timeout should be, since |
| 207 | libusb_get_next_timeout() will always return 0 and the timeout events will |
| 208 | be triggered as regular activity on the file descriptor set. |
| 209 | |
| 210 | Add API function libusb_pollfds_handle_timeouts() to detect whether |
| 211 | you're on a platform with the timing headache, and flesh out the |
| 212 | surrounding documentation. |
| 213 | |
| 214 | commit 9b120c2b3735566533c179aa8ca758fe45899a38 |
| 215 | Author: Daniel Drake <dsd@gentoo.org> |
| 216 | Date: Sat Nov 7 10:03:07 2009 +0000 |
| 217 | |
| 218 | Use AM_SILENT_RULES for building |
| 219 | |
| 220 | commit ef6ea6c3ae38e4524f10e16e8cb88177d39c4826 |
| 221 | Author: Daniel Drake <dsd@gentoo.org> |
| 222 | Date: Fri Nov 6 21:54:08 2009 +0000 |
| 223 | |
| 224 | v1.0.4 release |
| 225 | |
| 226 | commit 217f57617e0cff0d1bd6d726b243f04c6b179773 |
| 227 | Author: David Moore <dcm@acm.org> |
| 228 | Date: Fri Nov 6 21:37:25 2009 +0000 |
| 229 | |
| 230 | Linux: Add support for the new URB_BULK_CONTINUATION flag |
| 231 | |
| 232 | Add support for the new USBDEVFS_URB_BULK_CONTINUATION flag to libusb. |
| 233 | |
| 234 | This flag, which is expected to be available in usbfs starting with |
| 235 | kernel 2.6.32, allows the kernel to cancel multiple URBs upon receipt |
| 236 | of a short packet. This capability allows libusb to preserve data |
| 237 | integrity of large bulk transfers that are split into multiple URBs. |
| 238 | Without this support, these URBs must be canceled in userspace upon |
| 239 | receipt of a short packet, a race condition against future transfers |
| 240 | which might partially fill these canceled URBs. |
| 241 | |
| 242 | This patch automatically detects whether a supported kernel is present |
| 243 | and enables the use of the flag when possible. |
| 244 | |
| 245 | [dsd: tweaks to supported kernel detection, and some inline |
| 246 | documentation of this mechanism] |
| 247 | |
| 248 | commit 69830057547396f893f0d7b3125a05d016313b10 |
| 249 | Author: Daniel Drake <dsd@gentoo.org> |
| 250 | Date: Wed Oct 28 15:13:22 2009 +0545 |
| 251 | |
| 252 | Transfer locking |
| 253 | |
| 254 | At least on Linux, there were some possible races that could occur if |
| 255 | a transfer is cancelled from one thread while another thread is handling |
| 256 | an event for that transfer, or for if a transfer completes while it is |
| 257 | still being submitted from another thread, etc. |
| 258 | |
| 259 | On the global level, transfers could be submitted and cancelled at the |
| 260 | same time. |
| 261 | |
| 262 | Fix those issues with transfer-level locks. |
| 263 | |
| 264 | commit 98f1b30d24359cb3185051b8df9ebb663cc10369 |
| 265 | Author: Daniel Drake <dsd@gentoo.org> |
| 266 | Date: Mon Sep 14 08:01:24 2009 +0100 |
| 267 | |
| 268 | Clarify that timeout 0 means unlimited timeout |
| 269 | |
| 270 | commit 858684f0dd25921e09565034a88709dbf6f6c61b |
| 271 | Author: Daniel Drake <dsd@gentoo.org> |
| 272 | Date: Fri Sep 11 22:09:12 2009 +0100 |
| 273 | |
| 274 | Linux: more flexibility with monotonic clock |
| 275 | |
| 276 | Some users have reported that CLOCK_MONOTONIC does not work on their |
| 277 | systems - I suspect it is available on x86 but perhaps not some |
| 278 | of the more uncommon architectures. We should fall back on |
| 279 | CLOCK_REALTIME in these cases. |
| 280 | |
| 281 | Also, CLOCK_MONOTONIC_RAW seems even more monotonic, so we should use |
| 282 | that if it is available. |
| 283 | |
| 284 | We now test different clock IDs during initialization to find the |
| 285 | best one that works. |
| 286 | |
| 287 | commit fe0d8dce1ed704915d501e7da700440c78144211 |
| 288 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 289 | Date: Fri Sep 11 18:00:29 2009 +0100 |
| 290 | |
| 291 | Darwin: handle overflows |
| 292 | |
| 293 | commit f46716f42040986203fa6e873bfdabe1be2900ec |
| 294 | Author: Daniel Drake <dsd@gentoo.org> |
| 295 | Date: Thu Aug 27 21:14:54 2009 +0545 |
| 296 | |
| 297 | v1.0.3 release |
| 298 | |
| 299 | commit 5b489b8b2a5aba7b8b804e5af6d2628735548238 |
| 300 | Author: Toby Peterson <toby@macports.org> |
| 301 | Date: Sun Aug 23 10:04:59 2009 +0545 |
| 302 | |
| 303 | Darwin: 64-bit type fixes |
| 304 | |
| 305 | commit ad8ae04d0b52009af0b1180e005f7554d2bbb26c |
| 306 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 307 | Date: Sun Aug 23 10:02:55 2009 +0545 |
| 308 | |
| 309 | Darwin: fix crash when reading descriptors after close |
| 310 | |
| 311 | Fix a crash which occurs if the user does the following sequence on a |
| 312 | device: open, close, get_configuration_descriptor. |
| 313 | |
| 314 | commit 45ae2aecf5512dcff059b2a416534e81c6a00c88 |
| 315 | Author: Daniel Drake <dsd@gentoo.org> |
| 316 | Date: Sat Aug 1 13:55:15 2009 +0545 |
| 317 | |
| 318 | move bug info to bug tracker |
| 319 | |
| 320 | Protection needed: http://www.libusb.org/ticket/4 |
| 321 | Losing data: fixed in previous commit |
| 322 | |
| 323 | commit 126129e174062c2a27423817a459e5113f777789 |
| 324 | Author: Daniel Drake <dsd@gentoo.org> |
| 325 | Date: Thu Jul 9 22:09:04 2009 +0100 |
| 326 | |
| 327 | Linux: try harder not to lose any data |
| 328 | |
| 329 | We would previously lose any data that was present on a cancelled URB. |
| 330 | Work harder to make sure this doesn't happen. |
| 331 | |
| 332 | commit 0334ee642b47dfe1ca9db64b22e7702ea14b3f09 |
| 333 | Author: Daniel Drake <dsd@gentoo.org> |
| 334 | Date: Sun Jun 28 19:49:10 2009 +0100 |
| 335 | |
| 336 | Add libusb_get_max_iso_packet_size() |
| 337 | |
| 338 | As pointed out by Dennis Muhlestein, libusb_get_max_packet_size() |
| 339 | doesn't really do what the documentation might suggest because it |
| 340 | does not consider the number of transaction opportunities per |
| 341 | microframe. |
| 342 | |
| 343 | Add a new function to do what is useful for isochronous I/O. |
| 344 | |
| 345 | commit 615f18e64e96ae4ecc8e43d0de00933059a5209a |
| 346 | Author: Daniel Drake <dsd@gentoo.org> |
| 347 | Date: Sat Jun 20 22:33:21 2009 +0100 |
| 348 | |
| 349 | Linux: fix sending of zero length bulk packets |
| 350 | |
| 351 | Note that there are is a kernel bug preventing this from working |
| 352 | properly at the moment, even after this fix. |
| 353 | |
| 354 | commit 86f79fbf61c2018bdf009c7ebf92b38f3a16fd0c |
| 355 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 356 | Date: Fri Jun 19 22:18:44 2009 +0100 |
| 357 | |
| 358 | Darwin: Don't cancel transfers on timeout |
| 359 | |
| 360 | ...because the OS does this for us. |
| 361 | |
| 362 | commit d4bd9ed4de19a9c766f7a23eea8c852cdd61c18f |
| 363 | Author: Daniel Drake <dsd@gentoo.org> |
| 364 | Date: Sun Jun 14 19:23:21 2009 +0100 |
| 365 | |
| 366 | Don't terminate enums with commas |
| 367 | |
| 368 | g++ -pedantic doesn't like this |
| 369 | Reported by Eberhard Mattes |
| 370 | |
| 371 | commit 8be256082eb9bd8e243d89529f742926bb29a21b |
| 372 | Author: Daniel Drake <dsd@gentoo.org> |
| 373 | Date: Sat Jun 13 17:19:07 2009 +0100 |
| 374 | |
| 375 | v1.0.2 release |
| 376 | |
| 377 | commit 060e006e663fd59c281be29b71eb197e02b210e8 |
| 378 | Author: Daniel Drake <dsd@gentoo.org> |
| 379 | Date: Wed Jun 10 21:42:05 2009 +0100 |
| 380 | |
| 381 | Linux: fix config descriptor parsing on big-endian systems |
| 382 | |
| 383 | Multi-byte fields in the configuration descriptors that come back from |
| 384 | usbfs are always in bus endian format. |
| 385 | |
| 386 | Thanks to Joe Jezak for help investigating and fixing this. |
| 387 | |
| 388 | commit c4a905022f684da9a4a853eb9232a81a53df2652 |
| 389 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 390 | Date: Sun Jun 7 22:29:35 2009 +0100 |
| 391 | |
| 392 | Darwin: improve handling of disconnected devices |
| 393 | |
| 394 | commit 6b69f54451762ef590b9c938ab000c07cf9099a3 |
| 395 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 396 | Date: Sun Jun 7 22:26:37 2009 +0100 |
| 397 | |
| 398 | Darwin: fix parsing of config descriptors |
| 399 | |
| 400 | This was a confusion between configuration numbers and zero-based |
| 401 | configuration indexes. |
| 402 | |
| 403 | commit 2b3a9ffa776b383cb2dbc3c55e490e32e4c3c22b |
| 404 | Author: Daniel Drake <dsd@gentoo.org> |
| 405 | Date: Sun Jun 7 22:19:53 2009 +0100 |
| 406 | |
| 407 | Eliminate -Wsign-compare compiler warnings |
| 408 | |
| 409 | This was due to an API inconsistency which can be safely worked around. |
| 410 | Hopefully we'll remember to fix the API next time we come to break |
| 411 | things. |
| 412 | |
| 413 | commit cad5cb55c37137e94e35c74fdabfe42a5cbd229b |
| 414 | Author: David Moore <dcm@acm.org> |
| 415 | Date: Wed May 27 23:15:54 2009 -0700 |
| 416 | |
| 417 | Make synchronous transfer APIs robust against signal interruption |
| 418 | |
| 419 | libusb_control_transfer and libusb_bulk_transfer are designed to be |
| 420 | synchronous such that control is not returned until the transfer |
| 421 | definitively succeeds or fails. That assumption is violated if a signal |
| 422 | interrupts these functions because there is no way for the application |
| 423 | to continue waiting for the transfer without resubmitting it. This |
| 424 | patch changes these synchronous APIs so they do not abort in the case of |
| 425 | a signal interruption. |
| 426 | |
| 427 | Signed-off-by: David Moore <dcm@acm.org> |
| 428 | |
| 429 | commit b501795985a23109f176d296e7b544b4c6354528 |
| 430 | Author: Alex Vatchenko <alex@fabulatech.com> |
| 431 | Date: Thu May 28 15:58:50 2009 -0400 |
| 432 | |
| 433 | pre-gcc-3.4 compatibility |
| 434 | |
| 435 | The -fvisibility and -Wno-pointer-sign options are not available on |
| 436 | old GCC versions. |
| 437 | |
| 438 | commit 00bb2805e994887f0a754a825c3ce03d22393386 |
| 439 | Author: Daniel Drake <dsd@gentoo.org> |
| 440 | Date: Tue May 26 15:53:26 2009 -0400 |
| 441 | |
| 442 | Fix memory leak in config descriptor parsing |
| 443 | |
| 444 | Pointed out by Martin Koegler. |
| 445 | |
| 446 | commit 068ff5b8a83fec0a9a91c80535a25b89a9ae64e8 |
| 447 | Author: Daniel Drake <dsd@gentoo.org> |
| 448 | Date: Tue May 12 19:20:51 2009 -0400 |
| 449 | |
| 450 | Update AUTHORS |
| 451 | |
| 452 | commit aa24c04e4043e39674b59ff0d302b2365cd0078f |
| 453 | Author: Daniel Drake <dsd@gentoo.org> |
| 454 | Date: Tue May 12 19:03:37 2009 -0400 |
| 455 | |
| 456 | v1.0.1 release |
| 457 | |
| 458 | commit 44767677447fae4267131f99c591b14117486cac |
| 459 | Author: Daniel Drake <dsd@gentoo.org> |
| 460 | Date: Tue May 12 19:01:44 2009 -0400 |
| 461 | |
| 462 | Add BUGS file |
| 463 | |
| 464 | Probably missed a couple of outstanding issues |
| 465 | |
| 466 | commit e0365a6f22e0e7f330ac931df031cf6f17f453d1 |
| 467 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 468 | Date: Sun Mar 22 21:13:29 2009 -0400 |
| 469 | |
| 470 | Darwin: get_config_descriptor bugfixes |
| 471 | |
| 472 | commit 5fd0e8478240fece646a58a3c6114001a73be99f |
| 473 | Author: Daniel Drake <dsd@gentoo.org> |
| 474 | Date: Wed Feb 25 12:33:33 2009 -0300 |
| 475 | |
| 476 | Fix compilation of Darwin backend |
| 477 | |
| 478 | My fault. Reported by ihryamzik@gmail.com |
| 479 | |
| 480 | commit b49f6bf5c910d0fd694ecf165d7927673707bff9 |
| 481 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 482 | Date: Mon Feb 16 21:39:29 2009 -0300 |
| 483 | |
| 484 | Darwin backend |
| 485 | |
| 486 | commit d859158581e9a3250f36cdeeb8ea67cda04053bd |
| 487 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 488 | Date: Mon Feb 16 21:30:49 2009 -0300 |
| 489 | |
| 490 | Make endianness macros endian-independent |
| 491 | |
| 492 | Implementation suggested by David Moore. Needed for proper universal |
| 493 | code support on Darwin. |
| 494 | |
| 495 | commit 9196f58bdc8b9e967261df39865215faa5d39cfa |
| 496 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 497 | Date: Mon Feb 16 21:25:18 2009 -0300 |
| 498 | |
| 499 | Abstract clock reading into OS layer |
| 500 | |
| 501 | This will differ on Linux and Darwin, at least. |
| 502 | |
| 503 | [dsd: minor style tweaks] |
| 504 | |
| 505 | commit e91207860cac09b3afaafc4c14221b78d585c59a |
| 506 | Author: Daniel Drake <dsd@gentoo.org> |
| 507 | Date: Mon Feb 16 21:20:28 2009 -0300 |
| 508 | |
| 509 | Fix endianness in device descriptors |
| 510 | |
| 511 | Pointed out by Nathan Hjelm. |
| 512 | |
| 513 | commit ccf6d3d2f6acf9b9a1cb5e1f1b03f382ae509b17 |
| 514 | Author: Nathan Hjelm <hjelmn@mac.com> |
| 515 | Date: Mon Feb 16 21:13:52 2009 -0300 |
| 516 | |
| 517 | Only link with librt on Linux |
| 518 | |
| 519 | [dsd: tweak configure.ac change] |
| 520 | |
| 521 | commit 894539931e4b4bd85708fe98b956049243cd6fb0 |
| 522 | Author: Hans Ulrich Niedermann <hun@n-dimensional.de> |
| 523 | Date: Mon Jan 26 00:31:48 2009 +0100 |
| 524 | |
| 525 | API docs: describe libusb_transfer_cb_fn type |
| 526 | |
| 527 | Add some text describing the libusb_transfer_cb_fn function type |
| 528 | with the semantics I have gathered from reading other parts of |
| 529 | the API docs, referring to the proper section for more details. |
| 530 | |
| 531 | [dsd: tweaked the description slightly] |
| 532 | |
| 533 | commit abe34a2656f8f9f21e53603796c536585e6233ef |
| 534 | Author: Hans Ulrich Niedermann <hun@n-dimensional.de> |
| 535 | Date: Sun Jan 25 18:21:59 2009 +0100 |
| 536 | |
| 537 | Make empty array in struct compatible with C99 |
| 538 | |
| 539 | If the compiler is known to be running in C99 mode, |
| 540 | use "flexible array members" ("foo[]"). |
| 541 | |
| 542 | If the compiler is running in any other mode, continue |
| 543 | using the non-standard but widely common "foo[0]" syntax. |
| 544 | |
| 545 | commit 620075c7400764d9bb539b5c02065c45c2e8251e |
| 546 | Author: Hans Ulrich Niedermann <hun@n-dimensional.de> |
| 547 | Date: Sun Jan 25 18:20:45 2009 +0100 |
| 548 | |
| 549 | Avoid signedness errors in API docs example code |
| 550 | |
| 551 | commit c754ae294cfe96ec4738d6641137c9e6c56330c7 |
| 552 | Author: Daniel Drake <dsd@gentoo.org> |
| 553 | Date: Sat Jan 17 00:25:40 2009 +0000 |
| 554 | |
| 555 | Decrement poll() result when internal pipe has been handled |
| 556 | |
| 557 | When we receive data on the internal control pipe, we need to correctly |
| 558 | decrement the number of ready file descriptors before passing on the |
| 559 | remaining work to the OS implementation. |
| 560 | |
| 561 | commit b9ca960f2ba271d2b1a58e22b7c70464d69f6c8a |
| 562 | Author: David Moore <dcm@acm.org> |
| 563 | Date: Sun Jan 11 21:46:17 2009 -0800 |
| 564 | |
| 565 | Prevent transfer from being submitted twice or improperly canceled |
| 566 | |
| 567 | This ensures that tpriv->urbs and tpriv->iso_urbs are always set to NULL |
| 568 | whenever a transfer is not submitted. In this way, submit_*_transfer() |
| 569 | and cancel_*_transfer() can error check to ensure that the transfer is |
| 570 | in the correct state to be either submitted or canceled, preventing |
| 571 | potential memory leaks or double frees. |
| 572 | |
| 573 | Signed-off-by: David Moore <dcm@acm.org> |
| 574 | |
| 575 | commit d2a8ec2da8abcd8f4648ed118da16191011982dd |
| 576 | Author: David Moore <dcm@acm.org> |
| 577 | Date: Wed Jan 7 22:31:09 2009 -0800 |
| 578 | |
| 579 | Linux: Fix race condition in cancel_bulk_transfer() |
| 580 | |
| 581 | This fixes a race condition in cancel_bulk_transfer(). In the old |
| 582 | version, awaiting_reap and awaiting_discard are incremented in |
| 583 | cancel_bulk_transfer() and decremented in handle_bulk_completion(). |
| 584 | However, since these events may take place in two different threads, |
| 585 | these variables may reach zero before all URBs have been canceled, |
| 586 | triggered spurious callbacks and duplicate frees. |
| 587 | |
| 588 | This changes the logic to use a single variable "num_retired" to replace |
| 589 | both awaiting_reap and awaiting_discard. num_retired is incremented |
| 590 | only in handle_bulk_completion() and thus there is no race. The handler |
| 591 | will know that all URBs have been canceled when num_retired becomes |
| 592 | equal to num_urbs. |
| 593 | |
| 594 | This change also simplifies a great deal of the logic in both functions |
| 595 | and is a net reduction in the amount of code. |
| 596 | |
| 597 | Note that some variables such as "reap_action" probably need to still be |
| 598 | protected by a mutex, and this patch does not address that issue. |
| 599 | |
| 600 | Signed-off-by: David Moore <dcm@acm.org> |
| 601 | |
| 602 | commit 34b9eebe35d8167d43cffb6ad6175f6b2251b572 |
| 603 | Author: Daniel Drake <dsd@gentoo.org> |
| 604 | Date: Sat Dec 13 20:06:49 2008 +0000 |
| 605 | |
| 606 | v1.0.0 release |
| 607 | |
| 608 | It's here! |
| 609 | |
| 610 | commit b1d636dde418dc8fe6a8f037dea0d800e57c6b94 |
| 611 | Author: Mikhail Gusarov <dottedmag@dottedmag.ent> |
| 612 | Date: Wed Dec 10 15:24:05 2008 +0600 |
| 613 | |
| 614 | EOL-whitespace fixes |
| 615 | |
| 616 | Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.ent> |
| 617 | |
| 618 | commit 4a9ac382ba5149ae5b3d0962d6a840b5e0dc13aa |
| 619 | Author: Daniel Drake <dsd@gentoo.org> |
| 620 | Date: Tue Dec 9 21:43:37 2008 +0000 |
| 621 | |
| 622 | Linux: fix reading of active configuration |
| 623 | |
| 624 | Shannon Chuang pointed out that we only read 1 byte into an |
| 625 | uninitialized integer, and then return the whole thing. |
| 626 | |
| 627 | commit 81a7310d92461ba7b1d98aeff7c4c007fab07120 |
| 628 | Author: Daniel Drake <dsd@gentoo.org> |
| 629 | Date: Tue Dec 9 21:39:11 2008 +0000 |
| 630 | |
| 631 | Fix double free in descriptor parsing error path |
| 632 | |
| 633 | Pointed out by Shannon Chuang. |
| 634 | |
| 635 | commit 9b4b53453db56ba9c1d707a645bbe6c7a02a3c81 |
| 636 | Author: Daniel Drake <dsd@gentoo.org> |
| 637 | Date: Fri Nov 21 13:28:32 2008 +0000 |
| 638 | |
| 639 | Improvements to multi-threaded I/O system |
| 640 | |
| 641 | Documentation brushed up, and I realised that another function is |
| 642 | needed for tight event handling loops -- they must be able to check |
| 643 | if an open/close operation is trying to interrupt it. |
| 644 | |
| 645 | commit 0e5b0fcb77a90b8bd95ad23669da472af31ef069 |
| 646 | Author: Daniel Drake <dsd@gentoo.org> |
| 647 | Date: Fri Nov 21 11:00:47 2008 +0000 |
| 648 | |
| 649 | SourceForge website upload system changed |
| 650 | |
| 651 | commit a133875e6e2f52eff53ac03f505d59cbdd33e178 |
| 652 | Author: Daniel Drake <dsd@gentoo.org> |
| 653 | Date: Fri Nov 21 10:54:57 2008 +0000 |
| 654 | |
| 655 | v0.9.4 release |
| 656 | |
| 657 | commit c32aa662769b676ff3247778664fccc71fc427ec |
| 658 | Author: Daniel Drake <dsd@gentoo.org> |
| 659 | Date: Thu Nov 20 15:31:10 2008 +0000 |
| 660 | |
| 661 | Pause event handling while opening and closing devices |
| 662 | |
| 663 | Ludovic Rousseau found that crashes often occur if you close a device |
| 664 | while another thread is doing event handling. |
| 665 | |
| 666 | Fix this by adding an internal control pipe, which the close routines |
| 667 | use to interrupt the event handler and obtain the event handling lock, |
| 668 | ensuring that no other thread is handling events while the device is |
| 669 | closed. After the close completes, it signals all the event handlers |
| 670 | to start up again using the usual mechanism. |
| 671 | |
| 672 | Also modified libusb_open() to do a similar thing, so that event |
| 673 | handlers are interrupted in order to realise that a new poll fd has |
| 674 | appeared. |
| 675 | |
| 676 | commit 1d7cf3d0fa8698eae25097cbda1870be90ff6f5e |
| 677 | Author: Bastien Nocera <hadess@hadess.net> |
| 678 | Date: Thu Oct 30 14:35:23 2008 +0000 |
| 679 | |
| 680 | Add libusb_attach_kernel_driver() |
| 681 | |
| 682 | Add support for re-attaching a driver to a device under Linux. |
| 683 | |
| 684 | [dsd: fixed handling of return value, and added LIBUSB_ERROR_BUSY case] |
| 685 | |
| 686 | commit 914a4e70657c86b5094770aa2d898c978b1cdf41 |
| 687 | Author: Daniel Drake <dsd@gentoo.org> |
| 688 | Date: Sun Nov 2 21:45:54 2008 +0000 |
| 689 | |
| 690 | Linux: handle low-level transfer errors |
| 691 | |
| 692 | Handle more URB error status codes, thanks to Lou and Alan Stern. |
| 693 | |
| 694 | commit d25b566b3b8febafdda4211de724b4727dd4b7e0 |
| 695 | Author: Aurelien Jarno <aurelien@aurel32.net> |
| 696 | Date: Thu Sep 4 13:50:49 2008 +0200 |
| 697 | |
| 698 | Support for out-of-tree building |
| 699 | |
| 700 | Contrary to libusb 0.1, libusb 1.0 does not support out-of-tree building |
| 701 | for the documentation part. This patch fixes that by using a |
| 702 | doxygen.cfg.in file, which contains @top_srcdir@ to refer to the |
| 703 | location of the libusb source code. |
| 704 | |
| 705 | commit 8674c67ef78e1cf89db1fa584a4304f7c5ddcc5f |
| 706 | Author: Daniel Drake <dsd@gentoo.org> |
| 707 | Date: Sun Nov 2 15:00:39 2008 +0000 |
| 708 | |
| 709 | Document the logging style |
| 710 | |
| 711 | Suggested by Lou. |
| 712 | |
| 713 | commit 94936cbcfe3f02eb65c8b91e29896604316259d8 |
| 714 | Author: Daniel Drake <dsd@gentoo.org> |
| 715 | Date: Wed Aug 27 22:44:24 2008 -0500 |
| 716 | |
| 717 | Async I/O documentation touchups |
| 718 | |
| 719 | commit 2f8f1b1a900f5b9828e5e2ff93c2b26a44fd9de2 |
| 720 | Author: Daniel Drake <dsd@gentoo.org> |
| 721 | Date: Wed Aug 27 21:23:10 2008 -0500 |
| 722 | |
| 723 | Doc: fix typo in events lock section |
| 724 | |
| 725 | commit 046e681e4ac6b697e40458ee0f5baf3cd8763bb0 |
| 726 | Author: Daniel Drake <dsd@gentoo.org> |
| 727 | Date: Sat Aug 23 01:00:21 2008 -0500 |
| 728 | |
| 729 | v0.9.3 release |
| 730 | |
| 731 | commit fec6eaa4eecdf7e0b8299157a5dabef94417f193 |
| 732 | Author: Daniel Drake <dsd@gentoo.org> |
| 733 | Date: Sat Aug 23 00:57:44 2008 -0500 |
| 734 | |
| 735 | Reset internal transfer flags on submit |
| 736 | |
| 737 | This fixes a problem pointed out by Lou, where resubmitting a transfer |
| 738 | that previously timed out never timed out again, as if the timeout had |
| 739 | been set to 0. |
| 740 | |
| 741 | commit 81627bd38900ec9701ab69c141aa51a9abea1f60 |
| 742 | Author: Daniel Drake <dsd@gentoo.org> |
| 743 | Date: Sun Aug 10 17:09:48 2008 -0500 |
| 744 | |
| 745 | Add libusb_get_configuration prototype |
| 746 | |
| 747 | Pointed out by Lou |
| 748 | |
| 749 | commit 1631eaefcc7cdf7047c6b9225618a9aa1b3d87d1 |
| 750 | Author: Daniel Drake <dsd@gentoo.org> |
| 751 | Date: Sat Jul 19 17:51:28 2008 -0500 |
| 752 | |
| 753 | v0.9.2 release |
| 754 | |
| 755 | commit 0d9c40d7f3eee15ada624b5caab2cfb912144fdc |
| 756 | Author: Daniel Drake <dsd@gentoo.org> |
| 757 | Date: Wed Jul 16 09:22:12 2008 -0500 |
| 758 | |
| 759 | Linux: fix sysfs directory handling |
| 760 | |
| 761 | The length of the directory name varies, e.g. when you have nested hubs. |
| 762 | Use dynamic allocation to be able to deal with any length of name. |
| 763 | |
| 764 | commit 66c9847ad54b126ccafd4877292454ab57d0c115 |
| 765 | Author: David Engraf <dogge2k@gmx.de> |
| 766 | Date: Wed Jul 16 09:16:38 2008 -0500 |
| 767 | |
| 768 | Linux: correct usage of MAX_PATH |
| 769 | |
| 770 | Thanks to clarification from Artem Egorkine, MAX_PATH already includes |
| 771 | space for the trailing NULL. |
| 772 | |
| 773 | commit 6494f07a6d1b73ce786ae581c4ff691e39e61d19 |
| 774 | Author: Daniel Drake <dsd@gentoo.org> |
| 775 | Date: Tue Jul 8 21:15:54 2008 -0500 |
| 776 | |
| 777 | Linux: fix bulk/iso transfer double-cancellation |
| 778 | |
| 779 | Reset counters to 0 so that it's possible to cancel a transfer twice |
| 780 | without breaking things. Not sure that I want to support this properly, |
| 781 | but this makes it work at least. |
| 782 | |
| 783 | commit 9140f5b414e37a3714cde9c6d6c87c963338a83a |
| 784 | Author: Daniel Drake <dsd@gentoo.org> |
| 785 | Date: Sat Jun 28 21:28:44 2008 -0500 |
| 786 | |
| 787 | v0.9.1 release |
| 788 | |
| 789 | commit 83a029062c1d2bfb584d7a6ee94915583d37464c |
| 790 | Author: Daniel Drake <dsd@gentoo.org> |
| 791 | Date: Sat Jun 28 21:24:49 2008 -0500 |
| 792 | |
| 793 | Add missing GET_CONTEXT() |
| 794 | |
| 795 | commit fbad9a5426f0369394c88987355a8fb06a741ca1 |
| 796 | Author: Daniel Drake <dsd@gentoo.org> |
| 797 | Date: Sat Jun 28 21:18:41 2008 -0500 |
| 798 | |
| 799 | Linux: Compatibility with new sysfs descriptors file |
| 800 | |
| 801 | As of 2.6.26, the descriptors file now includes all descriptors, not |
| 802 | just the active one. |
| 803 | |
| 804 | commit 819e65f880ca43526036e56c65c415042c91f58f |
| 805 | Author: Daniel Drake <dsd@gentoo.org> |
| 806 | Date: Thu Jun 26 22:47:05 2008 -0500 |
| 807 | |
| 808 | Allow user data pointer to be passed through pollfd notification API |
| 809 | |
| 810 | commit 1df713d622ab4f0b03aad72d903ac7beb8fb3b90 |
| 811 | Author: Daniel Drake <dsd@gentoo.org> |
| 812 | Date: Tue Jun 24 23:01:51 2008 -0500 |
| 813 | |
| 814 | Introduce contexts to the API |
| 815 | |
| 816 | Suggested by David Zeuthen. This allows multiple libraries in the same |
| 817 | process to independently use libusb without interfering. |
| 818 | |
| 819 | commit 9818151c60a85aea6af24cb0996a92c3726c9864 |
| 820 | Author: Daniel Drake <dsd@gentoo.org> |
| 821 | Date: Tue Jun 24 21:31:13 2008 -0500 |
| 822 | |
| 823 | Revert "Temporary workaround for event handling serialization issue" |
| 824 | |
| 825 | This reverts commit 2d3a1111caff40ebb87983c861ff548cdc9e5946. |
| 826 | This was based on the assumption that dying threads would automatically |
| 827 | release mutexes, which is not the case. |
| 828 | |
| 829 | commit d5f82893fab3f1c13b1af4ba17aac72479bad7d5 |
| 830 | Author: Daniel Drake <dsd@gentoo.org> |
| 831 | Date: Fri Jun 20 23:04:53 2008 -0500 |
| 832 | |
| 833 | Overflow handling |
| 834 | |
| 835 | commit 546dee211eefbdd280fd1fc9dee84a9b52105078 |
| 836 | Author: Artem Egorkine <arteme@gmail.com> |
| 837 | Date: Tue Jun 17 18:27:38 2008 -0500 |
| 838 | |
| 839 | a fix for SIGSEGV in handle_bulk_completion() |
| 840 | |
| 841 | We cannot dereference tpriv after calling |
| 842 | usbi_handle_transfer_cancellation() because that function may invoke |
| 843 | the user-supplied callback which may free the transfer. |
| 844 | |
| 845 | commit 947ba8056456a5215724fb502e3e09d50016f699 |
| 846 | Author: Daniel Drake <dsd@gentoo.org> |
| 847 | Date: Mon Jun 16 22:50:50 2008 -0500 |
| 848 | |
| 849 | Refine configuration selection again |
| 850 | |
| 851 | At Alan Stern's suggestion, just offer the bare "set configuration" and |
| 852 | "get configuration" functionality, and let applications worry about the |
| 853 | specific race conditions and unusual situations. |
| 854 | |
| 855 | commit e7a7a49d0331ee0e14145f6e7ec39763b36314ac |
| 856 | Author: Daniel Drake <dsd@gentoo.org> |
| 857 | Date: Sun Jun 15 15:29:38 2008 -0500 |
| 858 | |
| 859 | docs update |
| 860 | |
| 861 | Add a THANKS file, make copyright notices easily accessible, update TODO |
| 862 | |
| 863 | commit 217534e4b396081e038ee5f94c813d1668963673 |
| 864 | Author: Daniel Drake <dsd@gentoo.org> |
| 865 | Date: Sun Jun 15 15:23:29 2008 -0500 |
| 866 | |
| 867 | round up poll timeout |
| 868 | |
| 869 | Pointed out by Richard Röjfors, otherwise we end up busy-looping with |
| 870 | a poll() timeout of 0. |
| 871 | |
| 872 | commit 5ad79b324bc5e11a75a585398a1e81c26f05e758 |
| 873 | Author: Felipe Balbi <me@felipebalbi.com> |
| 874 | Date: Sun Jun 8 12:34:48 2008 +0300 |
| 875 | |
| 876 | fix doc about libusb_open() |
| 877 | |
| 878 | libusb_open() returns int instead of libusb_device_handle. |
| 879 | |
| 880 | Signed-off-by: Felipe Balbi <me@felipebalbi.com> |
| 881 | [dsd: small correction] |
| 882 | |
| 883 | commit 2d3a1111caff40ebb87983c861ff548cdc9e5946 |
| 884 | Author: Daniel Drake <dsd@gentoo.org> |
| 885 | Date: Thu May 29 17:26:26 2008 +0100 |
| 886 | |
| 887 | Temporary workaround for event handling serialization issue |
| 888 | |
| 889 | Ludovic Rousseau pointed out that libusb_unlock_events() is not called |
| 890 | when a thread gets terminated with a signal, meaning that event waiters |
| 891 | will not be woken up in this case. |
| 892 | |
| 893 | Add a temporary hack to libusb_event_handler_active() so that at least |
| 894 | the other threads will realise on the next iteration of their event |
| 895 | handling loop. |
| 896 | |
| 897 | The real fix will likely involve reworking most of this. |
| 898 | |
| 899 | commit 514bb8790cfe8b93ccfff82bc17081b1030acce0 |
| 900 | Author: Daniel Drake <dsd@gentoo.org> |
| 901 | Date: Thu May 29 12:35:01 2008 +0100 |
| 902 | |
| 903 | Refine libusb_set_configuration() semantics |
| 904 | |
| 905 | Applications will generally want to set a configuration before claiming |
| 906 | interfaces. The problem is that the interface may already be set, and |
| 907 | someone else may have claimed an interface (meaning that all calls to |
| 908 | set_configuration will fail, even if it's for the same configuration). |
| 909 | |
| 910 | There are now 2 options: |
| 911 | 1. Use the new libusb_get_configuration() to determine active |
| 912 | configuration before calling libusb_set_configuration() |
| 913 | 2. Or just call libusb_set_configuration() as usual, which will do |
| 914 | nothing if that configuration is already active. |
| 915 | |
| 916 | commit d1292f8e7300051239a7ed2769d221dc7a6f9fca |
| 917 | Author: Daniel Drake <dsd@gentoo.org> |
| 918 | Date: Sun May 25 23:42:42 2008 +0100 |
| 919 | |
| 920 | v0.9.0 release |
| 921 | |
| 922 | First libusb-1.0 beta release |
| 923 | |
| 924 | commit 4cd249388e4ed2ca03f1263984ab67df5030bb65 |
| 925 | Author: Daniel Drake <dsd@gentoo.org> |
| 926 | Date: Sun May 25 22:59:09 2008 +0100 |
| 927 | |
| 928 | Take lock before raising event waiters condition |
| 929 | |
| 930 | This avoids a race between the user checking for active event handler |
| 931 | and then blocking on the condition variable |
| 932 | |
| 933 | commit 7c525480ab3c4db9205bfe30bb0d2bef9096c444 |
| 934 | Author: Daniel Drake <dsd@gentoo.org> |
| 935 | Date: Fri May 23 15:57:07 2008 +0100 |
| 936 | |
| 937 | Implement serialization of event handlers |
| 938 | |
| 939 | Now offers a mechanism to wait for events while another thread is |
| 940 | doing the event handling. Complicates things for MT async apps, but |
| 941 | then again it's a bit of a tricky combination to start with. |
| 942 | |
| 943 | commit ff0660a415ecfd0879600eaad1c5899b6d93a30a |
| 944 | Author: Daniel Drake <dsd@gentoo.org> |
| 945 | Date: Sat May 24 21:28:31 2008 +0100 |
| 946 | |
| 947 | allow LIBUSB_DEBUG env var to control message verbosity at runtime |
| 948 | |
| 949 | Based on ideas from Ludovic Rousseau |
| 950 | |
| 951 | commit 4d788967e3f8d75eaf3a1ac1ee8e2e8bed0601c1 |
| 952 | Author: Daniel Drake <dsd@gentoo.org> |
| 953 | Date: Fri May 23 15:58:11 2008 +0100 |
| 954 | |
| 955 | don't print messages by default |
| 956 | |
| 957 | Add libusb_set_debug() API to set message verbosity. |
| 958 | |
| 959 | Ludovic Rousseau pointed out that applications may close stdout/stderr |
| 960 | descriptors, which might then be reused. |
| 961 | |
| 962 | commit bef33bb9eba0da04ee7488d9cd5e6ab12bc61c0c |
| 963 | Author: Daniel Drake <dsd@gentoo.org> |
| 964 | Date: Mon May 19 15:43:27 2008 +0100 |
| 965 | |
| 966 | Fix memory leak in libusb_control_transfer |
| 967 | |
| 968 | Pointed out by David Engraf: we weren't freeing the transfer buffer |
| 969 | |
| 970 | commit a345bacb6f07ebb2122402041a9e8092d4a20a4d |
| 971 | Author: David Engraf <david.engraf@netcom.eu> |
| 972 | Date: Mon May 19 15:16:32 2008 +0100 |
| 973 | |
| 974 | Linux: fix handling of ioctl failure |
| 975 | |
| 976 | The return value of some ioctl commands in linux_usbfs.c are not |
| 977 | handeled correct. The ioctl function returns != 0 and errno is set with |
| 978 | the error code. |
| 979 | |
| 980 | commit e44396a458ecea9e5edd9a7577e617571c76860d |
| 981 | Author: David Engraf <david.engraf@netcom.eu> |
| 982 | Date: Mon May 19 15:13:38 2008 +0100 |
| 983 | |
| 984 | critical memory leak in handle_events |
| 985 | |
| 986 | This patch closes a critical memory leak in handle_events. The fds |
| 987 | variable is malloced but never freed. When I'm calling |
| 988 | handle_events with a timeout of 0, my system runs out of memory after a |
| 989 | few seconds. |
| 990 | |
| 991 | commit eb25630f52bc9848b444e439632c899977d887b0 |
| 992 | Author: Daniel Drake <dsd@gentoo.org> |
| 993 | Date: Fri May 16 23:50:20 2008 +0100 |
| 994 | |
| 995 | More informative libusb_open() return code |
| 996 | |
| 997 | Hopefully one of the last API tweaks... |
| 998 | |
| 999 | commit 2b2e9c40b195261b09ac52ebdb93eef25c79de90 |
| 1000 | Author: Daniel Drake <dsd@gentoo.org> |
| 1001 | Date: Fri May 16 22:37:56 2008 +0100 |
| 1002 | |
| 1003 | Fix endianness with descriptor handling |
| 1004 | |
| 1005 | Alan Stern pointed out that usbfs gives host-endian data, but sysfs gives |
| 1006 | bus-endian. |
| 1007 | |
| 1008 | commit 1298c51f516a7bf04ca9add1b7db14417cdc66f3 |
| 1009 | Author: Daniel Drake <dsd@gentoo.org> |
| 1010 | Date: Mon May 12 18:46:37 2008 +0100 |
| 1011 | |
| 1012 | Backend documentation for porting efforts |
| 1013 | |
| 1014 | Hopefully comprehensive enough for people to get started. |
| 1015 | |
| 1016 | commit ade26afc42c34ceb1c45afcadd2ea5e8240eaca4 |
| 1017 | Author: Daniel Drake <dsd@gentoo.org> |
| 1018 | Date: Mon May 12 15:43:30 2008 +0100 |
| 1019 | |
| 1020 | Linux: fix caching of guessed configuration |
| 1021 | |
| 1022 | Reported and tested by Xiaofan Chen |
| 1023 | |
| 1024 | commit e25d590a9198995b4f0b6afeb41ecae318715e7e |
| 1025 | Author: Rob Walker <rob@tenfoot.org.uk> |
| 1026 | Date: Sun May 11 21:14:18 2008 +0100 |
| 1027 | |
| 1028 | Linux: fix bulk transfer early completion |
| 1029 | |
| 1030 | We were forgetting about the remaining urbs when a non-final urb |
| 1031 | completed early. |
| 1032 | |
| 1033 | [dsd: some touchups and a warning message for a corner case that we don't handle] |
| 1034 | |
| 1035 | commit fec7c84163e25b8f811632828334d75da82bcb16 |
| 1036 | Author: Daniel Drake <dsd@gentoo.org> |
| 1037 | Date: Sun May 11 20:31:58 2008 +0100 |
| 1038 | |
| 1039 | Handle hot-unplugging |
| 1040 | |
| 1041 | This involved moving from select() to poll() because there is no way to |
| 1042 | distinguish usbfs's POLLERR condition with select(). |
| 1043 | |
| 1044 | commit aeb905fa5d78cdbba80a680aa7a2bb7338f27f65 |
| 1045 | Author: Daniel Drake <dsd@gentoo.org> |
| 1046 | Date: Sun May 11 16:04:35 2008 +0100 |
| 1047 | |
| 1048 | Documentation work |
| 1049 | |
| 1050 | commit ba5d9a45c06311204f51faef41d1ee215bb5b823 |
| 1051 | Author: Daniel Drake <dsd@gentoo.org> |
| 1052 | Date: Sun May 11 15:36:24 2008 +0100 |
| 1053 | |
| 1054 | Endianness of control setup packets |
| 1055 | |
| 1056 | Document behaviour where host-endian and bus-endian (little) should be |
| 1057 | used respectively. |
| 1058 | |
| 1059 | Also remove packed attribute from libusb_control_setup as all fields |
| 1060 | are naturally aligned. |
| 1061 | |
| 1062 | commit 7bedc3b2683f35ee1dd39a2ebe0ec05be0019f38 |
| 1063 | Author: Daniel Drake <dsd@gentoo.org> |
| 1064 | Date: Sun May 11 15:11:59 2008 +0100 |
| 1065 | |
| 1066 | Document that releasing interface causes SET_INTERFACE control request |
| 1067 | |
| 1068 | commit f2ede9876cd4f5cfa7751b975670fa449187fe3d |
| 1069 | Author: Daniel Drake <dsd@gentoo.org> |
| 1070 | Date: Sat May 10 21:45:42 2008 +0100 |
| 1071 | |
| 1072 | Support unconfigured devices |
| 1073 | |
| 1074 | commit 8ebb4ccdfaf5f095a1c38787d909d280ea64405c |
| 1075 | Author: Daniel Drake <dsd@gentoo.org> |
| 1076 | Date: Sat May 10 20:56:44 2008 +0100 |
| 1077 | |
| 1078 | Linux: fix caching of active configuration for non-sysfs |
| 1079 | |
| 1080 | commit c3844f7aeb2176636ce6e6ef697659fdb0b30048 |
| 1081 | Author: Daniel Drake <dsd@gentoo.org> |
| 1082 | Date: Sat May 10 14:42:43 2008 +0100 |
| 1083 | |
| 1084 | Fetch configurations by index (not value) |
| 1085 | |
| 1086 | Otherwise there is no way to know which values to look for. |
| 1087 | |
| 1088 | commit d77052c0d630e33737c38d601fd633155f6b2229 |
| 1089 | Author: Daniel Drake <dsd@gentoo.org> |
| 1090 | Date: Sat May 10 00:12:53 2008 +0100 |
| 1091 | |
| 1092 | Linux: comprehensive sysfs vs usbfs access |
| 1093 | |
| 1094 | Be more flexible when certain parts of sysfs are not available. |
| 1095 | |
| 1096 | commit 74bc842bac1a32a26323da6c3e8af2f66e1b5cfa |
| 1097 | Author: Daniel Drake <dsd@gentoo.org> |
| 1098 | Date: Fri May 9 19:38:13 2008 +0100 |
| 1099 | |
| 1100 | Linux: no need to store usbfs node path |
| 1101 | |
| 1102 | This can be computed from bus number and device address |
| 1103 | |
| 1104 | commit fe4adcc99e30115204ab832ad3e0170c9aca7629 |
| 1105 | Author: Daniel Drake <dsd@gentoo.org> |
| 1106 | Date: Fri May 9 14:34:31 2008 +0100 |
| 1107 | |
| 1108 | Rework configuration handling |
| 1109 | |
| 1110 | libusb no longer caches descriptors in libusb_device but backends are |
| 1111 | intended to be able to provide copies from memory. In the common linux |
| 1112 | case we can use sysfs. |
| 1113 | |
| 1114 | commit 5741bfe01a2481b8c3830c80edc3637bf62a7e16 |
| 1115 | Author: Daniel Drake <dsd@gentoo.org> |
| 1116 | Date: Fri May 9 14:36:14 2008 +0100 |
| 1117 | |
| 1118 | libusb_get_device_list() can return negative error code |
| 1119 | |
| 1120 | commit 45413101b78298e9332b22a34bc6bc159000ad8a |
| 1121 | Author: Daniel Drake <dsd@gentoo.org> |
| 1122 | Date: Thu May 8 23:09:21 2008 +0100 |
| 1123 | |
| 1124 | Linux: fix clear_halt implementation |
| 1125 | |
| 1126 | The ioctl expects an integer. |
| 1127 | Bug found with the assistance of Soumen Mondal |
| 1128 | |
| 1129 | commit 10d4e427cc171dfd6ad7f43a33ce3cfebcd7aa04 |
| 1130 | Author: Daniel Drake <dsd@gentoo.org> |
| 1131 | Date: Thu May 8 23:04:52 2008 +0100 |
| 1132 | |
| 1133 | Documentation work |
| 1134 | |
| 1135 | Partially based on some libusb-devel discussions |
| 1136 | |
| 1137 | commit ded0a249322571a075e3ed3528021864247dfa55 |
| 1138 | Author: Daniel Drake <dsd@gentoo.org> |
| 1139 | Date: Mon May 5 22:57:48 2008 +0100 |
| 1140 | |
| 1141 | Linux: enumerate devices and descriptors from sysfs |
| 1142 | |
| 1143 | Suggested by Alan Stern. This avoids waking up any suspended USB devices. |
| 1144 | sysfs is not available on all systems, so the usbfs mechanism is still |
| 1145 | in place as a fallback. |
| 1146 | |
| 1147 | commit 885c2a5de69d6b7d8902bb55d6d83680a5a1a6e5 |
| 1148 | Author: Daniel Drake <dsd@gentoo.org> |
| 1149 | Date: Mon May 5 21:34:31 2008 +0100 |
| 1150 | |
| 1151 | documentation touchups |
| 1152 | |
| 1153 | commit 59c205d542b43d79fe28622dbe8f03a3a3300b6f |
| 1154 | Author: Daniel Drake <dsd@gentoo.org> |
| 1155 | Date: Mon May 5 21:17:03 2008 +0100 |
| 1156 | |
| 1157 | more error code sanitization |
| 1158 | |
| 1159 | commit 88055d4b5913102a90ff666f75fd922c74860dc5 |
| 1160 | Author: Daniel Drake <dsd@gentoo.org> |
| 1161 | Date: Mon May 5 20:57:43 2008 +0100 |
| 1162 | |
| 1163 | Isochronous transfer helper functions |
| 1164 | |
| 1165 | commit a95c943ed301dcd20e92b0b3b255568899a4c42e |
| 1166 | Author: Daniel Drake <dsd@gentoo.org> |
| 1167 | Date: Mon May 5 17:54:33 2008 +0100 |
| 1168 | |
| 1169 | free open devices on exit |
| 1170 | |
| 1171 | commit b27fff633843824744df7d334cb89ece329cafa6 |
| 1172 | Author: Daniel Drake <dsd@gentoo.org> |
| 1173 | Date: Mon May 5 17:47:49 2008 +0100 |
| 1174 | |
| 1175 | move descriptor parsing into main library |
| 1176 | |
| 1177 | OS modules now provide functionality for fetching device/config |
| 1178 | descriptors |
| 1179 | |
| 1180 | commit a304eca71f22c6df7d70a901483b30f1b8e93378 |
| 1181 | Author: Daniel Drake <dsd@gentoo.org> |
| 1182 | Date: Mon May 5 16:22:33 2008 +0100 |
| 1183 | |
| 1184 | Detect endpoint halts and unsupported control requests |
| 1185 | |
| 1186 | commit 7da521954ba661d3f537440c31a84b66e974d56b |
| 1187 | Author: Daniel Drake <dsd@gentoo.org> |
| 1188 | Date: Mon May 5 01:00:31 2008 +0100 |
| 1189 | |
| 1190 | Linux: fix logical URB allocation |
| 1191 | |
| 1192 | We were allocating the wrong number of URBs for transfers of size |
| 1193 | multiples of 16k. |
| 1194 | |
| 1195 | commit 470b1bc42bf53373ce678fc76bab9160a54d6881 |
| 1196 | Author: Daniel Drake <dsd@gentoo.org> |
| 1197 | Date: Sun May 4 16:51:23 2008 +0100 |
| 1198 | |
| 1199 | add functionality for querying and detaching kernel driver |
| 1200 | |
| 1201 | commit 17ecfb0ecc833596c43755c80d461cddb9b3b0d7 |
| 1202 | Author: Daniel Drake <dsd@gentoo.org> |
| 1203 | Date: Sun May 4 16:20:46 2008 +0100 |
| 1204 | |
| 1205 | Descriptor reading functionality |
| 1206 | |
| 1207 | commit bdce367d1bd8691465844b2411c85215498f517d |
| 1208 | Author: Daniel Drake <dsd@gentoo.org> |
| 1209 | Date: Sun May 4 14:22:16 2008 +0100 |
| 1210 | |
| 1211 | sanitized error returns from synchronous I/O functions |
| 1212 | |
| 1213 | commit bfe74e9cd9c17a40fff042ea0647326f51cfecae |
| 1214 | Author: Daniel Drake <dsd@gentoo.org> |
| 1215 | Date: Sun May 4 14:10:19 2008 +0100 |
| 1216 | |
| 1217 | implement libusb_reset_device() |
| 1218 | |
| 1219 | commit c01f744ce07bbeccbe353e956479f0cc5a811a6f |
| 1220 | Author: Daniel Drake <dsd@gentoo.org> |
| 1221 | Date: Sun May 4 13:46:41 2008 +0100 |
| 1222 | |
| 1223 | Add functionality to clear endpoint halt conditions |
| 1224 | |
| 1225 | commit 5878daa85e3364bb3885190425d4f9deaa2d8c36 |
| 1226 | Author: Daniel Drake <dsd@gentoo.org> |
| 1227 | Date: Sun May 4 00:51:59 2008 +0100 |
| 1228 | |
| 1229 | Fix libusb_get_device_list return value |
| 1230 | |
| 1231 | It was documented to return the list length, but was returning 0. |
| 1232 | |
| 1233 | commit a74106a9b48b88bae0e099be34a18d653a8f1238 |
| 1234 | Author: Daniel Drake <dsd@gentoo.org> |
| 1235 | Date: Sat May 3 16:50:58 2008 +0100 |
| 1236 | |
| 1237 | Well-defined error codes for some operations |
| 1238 | |
| 1239 | Also renamed set_interface_altsetting to set_interface_alt_setting for |
| 1240 | better consistency |
| 1241 | |
| 1242 | commit 0f463fe671455670efcf4a93e526b2a9082b0afe |
| 1243 | Author: Daniel Drake <dsd@gentoo.org> |
| 1244 | Date: Sat May 3 16:24:11 2008 +0100 |
| 1245 | |
| 1246 | Add libusb_set_configuration() |
| 1247 | |
| 1248 | commit e65608994f795f97e4d6adac20329ca08709db2d |
| 1249 | Author: Daniel Drake <dsd@gentoo.org> |
| 1250 | Date: Sat May 3 15:40:23 2008 +0100 |
| 1251 | |
| 1252 | rename LIBUSB_ERROR_NOMEM to LIBUSB_ERROR_NO_MEM |
| 1253 | improves consistency |
| 1254 | |
| 1255 | commit 5384bbf300779e7936eac9ebd33ee38f5b533f8e |
| 1256 | Author: Daniel Drake <dsd@gentoo.org> |
| 1257 | Date: Sat May 3 14:34:54 2008 +0100 |
| 1258 | |
| 1259 | Linux: better EACCES handling |
| 1260 | |
| 1261 | commit 1936eeeba3f933a078df1005cc7c1533a8ec2bac |
| 1262 | Author: Daniel Drake <dsd@gentoo.org> |
| 1263 | Date: Sat May 3 14:26:56 2008 +0100 |
| 1264 | |
| 1265 | Rename reference counting API |
| 1266 | |
| 1267 | This improves consistency with the naming style of other functions |
| 1268 | |
| 1269 | commit e2f822b4f17342f614ab2455179a80975fca6cb0 |
| 1270 | Author: Daniel Drake <dsd@gentoo.org> |
| 1271 | Date: Sat May 3 14:24:40 2008 +0100 |
| 1272 | |
| 1273 | Linux: use read-only access to enumerate devices and descriptors |
| 1274 | |
| 1275 | Write access is not needed until later. |
| 1276 | Also fix handling of open() error code. |
| 1277 | |
| 1278 | commit 81204eed3d0a7744b89b68bd90d567b68d191a91 |
| 1279 | Author: Daniel Drake <dsd@gentoo.org> |
| 1280 | Date: Fri May 2 17:45:23 2008 +0100 |
| 1281 | |
| 1282 | document pollfd notifiers |
| 1283 | |
| 1284 | commit e20f8281fb7da32587f81b5eea5818af1eab0fe0 |
| 1285 | Author: Daniel Drake <dsd@gentoo.org> |
| 1286 | Date: Fri May 2 17:33:52 2008 +0100 |
| 1287 | |
| 1288 | Allow multiple interfaces to be claimed |
| 1289 | |
| 1290 | commit 66685ff25d81a1f1a2a3dbda882d407638d00672 |
| 1291 | Author: Daniel Drake <dsd@gentoo.org> |
| 1292 | Date: Fri May 2 16:55:36 2008 +0100 |
| 1293 | |
| 1294 | Documentation improvements |
| 1295 | |
| 1296 | commit 55bce39cbe0c9746a428b66751d1afe551d86381 |
| 1297 | Author: Daniel Drake <dsd@gentoo.org> |
| 1298 | Date: Fri May 2 12:55:34 2008 +0100 |
| 1299 | |
| 1300 | detect short descriptor reads |
| 1301 | |
| 1302 | commit f3f17846dcced6b2dc40f899ed6a2aba151c264b |
| 1303 | Author: Daniel Drake <dsd@gentoo.org> |
| 1304 | Date: Fri May 2 12:32:46 2008 +0100 |
| 1305 | |
| 1306 | remove "references" and "referenced by" listings in documentation |
| 1307 | |
| 1308 | commit fbb017dd9a26ce449578482b95157b3c850fa446 |
| 1309 | Author: Daniel Drake <dsd@gentoo.org> |
| 1310 | Date: Fri May 2 12:32:08 2008 +0100 |
| 1311 | |
| 1312 | Use typedef ptr types in parameters and return types |
| 1313 | |
| 1314 | This is passed through to the docs |
| 1315 | |
| 1316 | commit 3675e978fb7a1042f8601931255658bcd14a2298 |
| 1317 | Author: Daniel Drake <dsd@gentoo.org> |
| 1318 | Date: Fri May 2 12:28:31 2008 +0100 |
| 1319 | |
| 1320 | Expose bus number and device address |
| 1321 | |
| 1322 | commit 2012898ffc492317cb428d97077bf5ee0e98e1da |
| 1323 | Author: Daniel Drake <dsd@gentoo.org> |
| 1324 | Date: Tue Apr 29 15:26:46 2008 +0100 |
| 1325 | |
| 1326 | standardised error codes |
| 1327 | |
| 1328 | commit 0269c75ae3de7acb2dc555e45928750e35fdb320 |
| 1329 | Author: Daniel Drake <dsd@gentoo.org> |
| 1330 | Date: Tue Apr 29 13:34:26 2008 +0100 |
| 1331 | |
| 1332 | Fix descriptor handling memory leaks |
| 1333 | |
| 1334 | commit d3ab4e3bd4068cba41c1e12c4b175018dc3cb343 |
| 1335 | Author: Daniel Drake <dsd@gentoo.org> |
| 1336 | Date: Sun Apr 27 23:50:01 2008 +0100 |
| 1337 | |
| 1338 | Rename endpoint_type to transfer_type |
| 1339 | |
| 1340 | Matches the USB specs more closely |
| 1341 | |
| 1342 | commit 379e90e0fdcef6f4db1849fa0a7fd524407cf564 |
| 1343 | Author: Daniel Drake <dsd@gentoo.org> |
| 1344 | Date: Sun Apr 27 23:41:24 2008 +0100 |
| 1345 | |
| 1346 | Remove a FIXME comment |
| 1347 | |
| 1348 | it can't. |
| 1349 | |
| 1350 | commit 2ddf81420557d016bbcbf28dda0d8d1c64d16264 |
| 1351 | Author: Daniel Drake <dsd@gentoo.org> |
| 1352 | Date: Sun Apr 27 23:30:34 2008 +0100 |
| 1353 | |
| 1354 | Fix unlocked flying_transfers list access |
| 1355 | |
| 1356 | commit 7ac0a405b4c57db42e88cbcba5f135697f03b646 |
| 1357 | Author: Daniel Drake <dsd@gentoo.org> |
| 1358 | Date: Sun Apr 27 23:27:04 2008 +0100 |
| 1359 | |
| 1360 | Remove libusb_cancel_transfer_sync |
| 1361 | |
| 1362 | This wasn't a particularly nice API. Cancellation should be handled by |
| 1363 | the completion handler. |
| 1364 | |
| 1365 | commit a5006fd7b9dae6f1db7969c8744086ba3a7c027b |
| 1366 | Author: Daniel Drake <dsd@gentoo.org> |
| 1367 | Date: Sun Apr 27 19:56:02 2008 +0100 |
| 1368 | |
| 1369 | remove extra timeout check |
| 1370 | |
| 1371 | If select() didn't timeout then don't check for more timeouts |
| 1372 | |
| 1373 | commit 3bdafaa359d9c65e1af360b5338e1f25528a8066 |
| 1374 | Author: Daniel Drake <dsd@gentoo.org> |
| 1375 | Date: Sun Apr 27 19:53:51 2008 +0100 |
| 1376 | |
| 1377 | Documentation touchups |
| 1378 | |
| 1379 | Detail that interrupt polling interval is fixed to endpoint bInterval, |
| 1380 | and that data may be transferred even on a timeout condition. |
| 1381 | |
| 1382 | commit e12618c79d520955b1abf80a277e881d91907162 |
| 1383 | Author: Daniel Drake <dsd@gentoo.org> |
| 1384 | Date: Sat Apr 19 16:17:53 2008 +0100 |
| 1385 | |
| 1386 | add threaded example program |
| 1387 | |
| 1388 | commit 1fcdb0678b759569db7cd530457dbc0a5f86fb1d |
| 1389 | Author: Daniel Drake <dsd@gentoo.org> |
| 1390 | Date: Fri Apr 18 23:58:54 2008 +0100 |
| 1391 | |
| 1392 | libusb_device mutex protection |
| 1393 | |
| 1394 | commit 0efd2efa65d5513e5754d717d522b2c5c45332e2 |
| 1395 | Author: Daniel Drake <dsd@gentoo.org> |
| 1396 | Date: Fri Apr 18 23:42:40 2008 +0100 |
| 1397 | |
| 1398 | mutex protection for device and handle lists |
| 1399 | |
| 1400 | commit d2ff2b14a15e5099ff0214502944de30f141a39c |
| 1401 | Author: Daniel Drake <dsd@gentoo.org> |
| 1402 | Date: Fri Apr 18 20:00:31 2008 +0100 |
| 1403 | |
| 1404 | lock flying_transfers during transfer removal |
| 1405 | |
| 1406 | commit 2569980762cb66804da31e54bef243702da8dfe9 |
| 1407 | Author: Daniel Drake <dsd@gentoo.org> |
| 1408 | Date: Fri Apr 18 19:59:56 2008 +0100 |
| 1409 | |
| 1410 | pollfds mutex protection |
| 1411 | |
| 1412 | commit f5a33e08ec0da51ccb5e95a5adec50c5211a9bf5 |
| 1413 | Author: Daniel Drake <dsd@gentoo.org> |
| 1414 | Date: Thu Apr 17 18:24:54 2008 +0100 |
| 1415 | |
| 1416 | flying_transfers mutex protection |
| 1417 | |
| 1418 | This marks the beginnings of thread safety |
| 1419 | |
| 1420 | commit 85e1994123545125ce0653415d122f9790f8d041 |
| 1421 | Author: Daniel Drake <dsd@gentoo.org> |
| 1422 | Date: Thu Apr 17 17:44:28 2008 +0100 |
| 1423 | |
| 1424 | Linux: fix control transfer cancellation |
| 1425 | |
| 1426 | commit cf9b831f7bfd97ba88721c1d934aeda522f3692c |
| 1427 | Author: Daniel Drake <dsd@gentoo.org> |
| 1428 | Date: Mon Apr 14 20:03:27 2008 +0100 |
| 1429 | |
| 1430 | set correct endpoint type for isochronous transfers |
| 1431 | |
| 1432 | Pointed out by Felix Domke |
| 1433 | |
| 1434 | commit ad6e2b712c5b54af44424e58a2776686314e26b7 |
| 1435 | Author: Daniel Drake <dsd@gentoo.org> |
| 1436 | Date: Fri Apr 11 21:40:07 2008 +0100 |
| 1437 | |
| 1438 | Linux: fire multiple URBs at once for split transfers |
| 1439 | |
| 1440 | This results in a significant performance increase for bulk transfers |
| 1441 | larger than 16kb, and fixes a bug where data would be spliced and mixed |
| 1442 | between two simultaneously submitted transfers to the same endpoint. |
| 1443 | |
| 1444 | It also allows isochronous transfers larger than 16kb to be submitted. |
| 1445 | |
| 1446 | This commit also improves cancellation - the library now understands |
| 1447 | what is going on. |
| 1448 | |
| 1449 | commit d9b2ce2894ed43d34de6850f2dac50ccabf1db55 |
| 1450 | Author: Daniel Drake <dsd@gentoo.org> |
| 1451 | Date: Tue Apr 1 23:28:32 2008 +0100 |
| 1452 | |
| 1453 | Constify some return data |
| 1454 | |
| 1455 | Make it clear that apps should not mess with these structures |
| 1456 | |
| 1457 | commit 8d809854e2b19c2b7c27ab05e5d76b34e2a5cead |
| 1458 | Author: Daniel Drake <dsd@gentoo.org> |
| 1459 | Date: Sun Mar 30 22:19:00 2008 +0100 |
| 1460 | |
| 1461 | Rename libusb_poll() to libusb_handle_events() |
| 1462 | |
| 1463 | commit 211f80c9f2a4a58cd2bbf5b7751f45089c8961e7 |
| 1464 | Author: Daniel Drake <dsd@gentoo.org> |
| 1465 | Date: Tue Mar 25 16:24:30 2008 +0000 |
| 1466 | |
| 1467 | Isochronous endpoint I/O |
| 1468 | |
| 1469 | Due to variable-sized structures, this involved changing allocation |
| 1470 | mechanism. All transfers must now be allocated and freed through |
| 1471 | libusb. |
| 1472 | |
| 1473 | A synchronous function is missing, and I could do with writing a few |
| 1474 | more helper functions to simplify things. |
| 1475 | |
| 1476 | commit b1ade6fca668d8aa156d5b5bf3a933f116144dc2 |
| 1477 | Author: Daniel Drake <dsd@gentoo.org> |
| 1478 | Date: Sun Mar 23 21:17:32 2008 +0000 |
| 1479 | |
| 1480 | distcheck fix for docs |
| 1481 | |
| 1482 | commit 62077b7fa5b222e1bb9dea3467585aed69d596e2 |
| 1483 | Author: Daniel Drake <dsd@gentoo.org> |
| 1484 | Date: Sun Mar 23 21:13:14 2008 +0000 |
| 1485 | |
| 1486 | Support for changing altsetting |
| 1487 | |
| 1488 | Will probably be suject to later consideration w.r.t. claiming of endpoints |
| 1489 | |
| 1490 | commit 0499e9f418607b5786ac0c3e97bc46737a609ca3 |
| 1491 | Author: Daniel Drake <dsd@gentoo.org> |
| 1492 | Date: Thu Mar 20 21:10:01 2008 +0000 |
| 1493 | |
| 1494 | it's called bmRequestType |
| 1495 | |
| 1496 | commit 88884ae4bbe721dac117d5190303cb67852fa5d3 |
| 1497 | Author: Daniel Drake <dsd@gentoo.org> |
| 1498 | Date: Thu Mar 20 21:07:00 2008 +0000 |
| 1499 | |
| 1500 | Save transfer flags before invoking callback |
| 1501 | |
| 1502 | The callback may free the transfer, in which case we can't read the flags |
| 1503 | after it has returned. |
| 1504 | |
| 1505 | commit ead09cde6895df0034a685516987ce253575e9a7 |
| 1506 | Author: Daniel Drake <dsd@gentoo.org> |
| 1507 | Date: Sat Mar 15 16:35:12 2008 +0000 |
| 1508 | |
| 1509 | API documentation |
| 1510 | |
| 1511 | Hopefully mostly complete. Some constants were renamed and move into |
| 1512 | enums. |
| 1513 | |
| 1514 | commit 285724cc14ea5f993e2c4c92fe0aaf6c335bc139 |
| 1515 | Author: Vasily Khoruzhick <anarsoul@gmail.com> |
| 1516 | Date: Sat Mar 15 13:50:53 2008 +0200 |
| 1517 | |
| 1518 | Fix get_device_list for realloc case |
| 1519 | |
| 1520 | get_device_list can modify pointer passed to it with realloc, but this |
| 1521 | case wasn't handled and caused crash on my machine. Fixed |
| 1522 | |
| 1523 | Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com> |
| 1524 | |
| 1525 | commit c0c9432d38b22784070dce3a7874c62c31786a27 |
| 1526 | Author: Daniel Drake <dsd@gentoo.org> |
| 1527 | Date: Thu Mar 13 12:36:56 2008 +0000 |
| 1528 | |
| 1529 | Beginnings of cross-platform abstraction |
| 1530 | |
| 1531 | This also includes a libusb_get_pollfds API change |
| 1532 | |
| 1533 | commit de4c5341d168697baa4c0901c406deb47e78aae7 |
| 1534 | Author: Daniel Drake <dsd@gentoo.org> |
| 1535 | Date: Mon Mar 10 18:38:15 2008 +0000 |
| 1536 | |
| 1537 | Add transfer flags |
| 1538 | |
| 1539 | commit aae05f688dc26a013166ab9436fd25018b49d76a |
| 1540 | Author: Daniel Drake <dsd@gentoo.org> |
| 1541 | Date: Mon Mar 10 11:32:15 2008 +0000 |
| 1542 | |
| 1543 | Simplify libusb_fill_control_transfer() |
| 1544 | |
| 1545 | Transfer length can be inferred from the setup packet. |
| 1546 | |
| 1547 | commit 7c5ea95297c2f3f6afc42f60c1bf1bef777bb1fb |
| 1548 | Author: Daniel Drake <dsd@gentoo.org> |
| 1549 | Date: Mon Mar 10 11:29:24 2008 +0000 |
| 1550 | |
| 1551 | get_device fixup |
| 1552 | |
| 1553 | Rename API away from "devh" and fix implementation to match header |
| 1554 | |
| 1555 | commit 9e89d1aa82cbe7e21672ee3e2e545238fffc5318 |
| 1556 | Author: Daniel Drake <dsd@gentoo.org> |
| 1557 | Date: Sun Mar 9 16:27:06 2008 +0000 |
| 1558 | |
| 1559 | Rename descriptor-getting functions |
| 1560 | |
| 1561 | Based on feedback from Tim Roberts |
| 1562 | |
| 1563 | commit 283ae96c3ccbcee1c3950fce18ae47093ae8200e |
| 1564 | Author: Daniel Drake <dsd@gentoo.org> |
| 1565 | Date: Sun Mar 9 16:15:02 2008 +0000 |
| 1566 | |
| 1567 | Simplify cancellation API |
| 1568 | |
| 1569 | No need to provide the device here. |
| 1570 | |
| 1571 | commit ebad1c79688ba603ad017ed0fe2f3c0dc8edd1ad |
| 1572 | Author: Daniel Drake <dsd@gentoo.org> |
| 1573 | Date: Sun Mar 9 16:12:08 2008 +0000 |
| 1574 | |
| 1575 | Rename libusb_dev_handle to libusb_device_handle |
| 1576 | |
| 1577 | commit 1ac0a7d88f282b6f293c456fac8edb143cbaca3d |
| 1578 | Author: Daniel Drake <dsd@gentoo.org> |
| 1579 | Date: Sun Mar 9 01:01:57 2008 +0000 |
| 1580 | |
| 1581 | Move synchronous I/O implementation to its own file |
| 1582 | |
| 1583 | commit 66348c90ea4570bf999ac301089e006d0cce1926 |
| 1584 | Author: Daniel Drake <dsd@gentoo.org> |
| 1585 | Date: Sun Mar 9 00:58:09 2008 +0000 |
| 1586 | |
| 1587 | Separate transfer allocation and submission |
| 1588 | |
| 1589 | Leads to some hefty API changes. Now we're much more similar to the |
| 1590 | Linux kernel model. |
| 1591 | |
| 1592 | Problems with dealing with asynchronous control transfers are passed |
| 1593 | on to the user, basically you must allocate a buffer, start with the |
| 1594 | setup, and put the data after. This won't make much sense until |
| 1595 | documented (soon...) |
| 1596 | |
| 1597 | commit d21ebe47ce578c93cd8969be1c933d503e32e5d4 |
| 1598 | Author: Daniel Drake <dsd@gentoo.org> |
| 1599 | Date: Sat Mar 8 12:48:35 2008 +0000 |
| 1600 | |
| 1601 | Rework URB API naming |
| 1602 | |
| 1603 | Now refer to everything as "transfers" as consistent with the USB spec |
| 1604 | libusb_transfer is now a kind of transfer handle. To reduce confusion |
| 1605 | with libusb_bulk_transfer and libusb_control_transfer, those have been |
| 1606 | renamed to libusb_{control,bulk}_transfer_request. |
| 1607 | |
| 1608 | commit fd6fb3cc093ad47271353a0e468d16547f991382 |
| 1609 | Author: Daniel Drake <dsd@gentoo.org> |
| 1610 | Date: Sat Mar 8 11:53:33 2008 +0000 |
| 1611 | |
| 1612 | Add fallback on /proc/bus/usb |
| 1613 | |
| 1614 | /dev/bus/usb is a relatively new thing probably not present on every |
| 1615 | system |
| 1616 | |
| 1617 | commit f3fdf447916289cd92b7190377681894e8ab611a |
| 1618 | Author: Daniel Drake <dsd@gentoo.org> |
| 1619 | Date: Sat Mar 8 00:10:16 2008 +0000 |
| 1620 | |
| 1621 | Persistent libusb_device storage |
| 1622 | |
| 1623 | Devices are now assigned a session ID (currently busnum:devaddr) which |
| 1624 | is used to distinguish unique devices. |
| 1625 | |
| 1626 | Now multiple callers of libusb_get_device_list will get the same |
| 1627 | libusb_device structure instances. |
| 1628 | |
| 1629 | commit 23f8fb8bafeeda93ce234cb40eb7219d2e36c19c |
| 1630 | Author: Daniel Drake <dsd@gentoo.org> |
| 1631 | Date: Thu Mar 6 23:43:57 2008 +0000 |
| 1632 | |
| 1633 | Add convenience function to find and open a device by USB VID+PID |
| 1634 | |
| 1635 | Lots of libusb apps I write are simple test apps not intended to be real |
| 1636 | apps. Having a function available to quickly locate my device will be |
| 1637 | handy in such situations. |
| 1638 | |
| 1639 | commit 9cfdb494fccac53a4277da7c8b6d15f1a72a4959 |
| 1640 | Author: Daniel Drake <dsd@gentoo.org> |
| 1641 | Date: Thu Mar 6 23:25:20 2008 +0000 |
| 1642 | |
| 1643 | Rework device discovery API |
| 1644 | |
| 1645 | libusb_find_devices and libusb_get_devices are no more |
| 1646 | |
| 1647 | libusb_get_device_list obtains a list of libusb_device structures for all |
| 1648 | known devices in the system. |
| 1649 | |
| 1650 | Each libusb_device now has a reference count, defaulting to 1 on |
| 1651 | instantiation. The reference count of 1 refers to the fact that it is |
| 1652 | present in the list in this scenario. |
| 1653 | |
| 1654 | Opening a device adds a pointer to the libusb_device structure in the |
| 1655 | handle, so that also adds a reference. Closing the device removes that |
| 1656 | reference. |
| 1657 | |
| 1658 | The function to free the device list can optionally unref all the devices |
| 1659 | inside. |
| 1660 | |
| 1661 | In future we will make the libusb_device instances all "global" so that if |
| 1662 | the app calls get_device_list twice it actually gets the same libusb_device |
| 1663 | structure references back. This way we can start to track disconnects, and |
| 1664 | we can investigate adding a unique "session ID" to each libusb_device, an |
| 1665 | identifier guaranteed to be unique to that device until reboot. |
| 1666 | |
| 1667 | commit 77cea822788e024c848c7f554915f771a2dc1e0f |
| 1668 | Author: Daniel Drake <dsd@gentoo.org> |
| 1669 | Date: Thu Mar 6 17:05:04 2008 +0000 |
| 1670 | |
| 1671 | Remove ctrl debug code |
| 1672 | |
| 1673 | This was unintentionally committed |
| 1674 | |
| 1675 | commit b873521a6f575628d535f89278bb67888004d0ca |
| 1676 | Author: Daniel Drake <dsd@gentoo.org> |
| 1677 | Date: Tue Feb 26 22:11:34 2008 +0000 |
| 1678 | |
| 1679 | Notifications for changes to the fd set |
| 1680 | |
| 1681 | Applications can now be notified when they should start and stop polling |
| 1682 | new file descriptors. |
| 1683 | |
| 1684 | commit ca66e16a370c828d08a5bc3f9f203db145c81a4c |
| 1685 | Author: Daniel Drake <dsd@gentoo.org> |
| 1686 | Date: Fri Feb 1 12:20:20 2008 +0000 |
| 1687 | |
| 1688 | Fix header installation path |
| 1689 | |
| 1690 | commit e583100a4086c411093eef5c450880b965df6b1b |
| 1691 | Author: Daniel Drake <dsd@gentoo.org> |
| 1692 | Date: Wed Jan 30 13:24:54 2008 +0000 |
| 1693 | |
| 1694 | TODO update |
| 1695 | |
| 1696 | commit 637a8d7ff8a11a23588925d9d3003a609bda8075 |
| 1697 | Author: Daniel Drake <dsd@gentoo.org> |
| 1698 | Date: Wed Jan 30 13:19:10 2008 +0000 |
| 1699 | |
| 1700 | Remove timers and signalfd |
| 1701 | |
| 1702 | Instead of timers, add a mechanism for informing the parent app when the |
| 1703 | next timeout is due to happen, so that it can call us at that time. |
| 1704 | |
| 1705 | As we no longer use signals, signalfd has also been removed. |
| 1706 | |
| 1707 | commit 8c56abe2e13ed4dd8c15737c21be7b6b1e5cf2ac |
| 1708 | Author: Daniel Drake <dsd@gentoo.org> |
| 1709 | Date: Sat Jan 26 18:16:56 2008 +0000 |
| 1710 | |
| 1711 | Poll usbfs descriptors directly |
| 1712 | |
| 1713 | We don't need to use signalfd for this functionality because we can poll |
| 1714 | the file descriptors for write events. |
| 1715 | |
| 1716 | commit e3ce855acec33d37dd833bca5dad5e8c3773003a |
| 1717 | Author: Daniel Drake <dsd@gentoo.org> |
| 1718 | Date: Sat Jan 26 17:26:40 2008 +0000 |
| 1719 | |
| 1720 | Fix dpfp example compilation |
| 1721 | |
| 1722 | commit 1d41a5cda2fefe826877ee4445199af30aa188b7 |
| 1723 | Author: Daniel Drake <dsd@gentoo.org> |
| 1724 | Date: Sat Jan 26 17:22:53 2008 +0000 |
| 1725 | |
| 1726 | C++ header safety |
| 1727 | |
| 1728 | Pointed out by JF |
| 1729 | |
| 1730 | commit 37b7f9a112eb1441d352467f661205b8872d0553 |
| 1731 | Author: Daniel Drake <dsd@gentoo.org> |
| 1732 | Date: Sat Jan 26 14:03:43 2008 +0000 |
| 1733 | |
| 1734 | Only use -fgnu89-inline when supported |
| 1735 | |
| 1736 | J F pointed out that this option only works on gcc 4.3 and newer. |
| 1737 | |
| 1738 | commit e3a09ca0b9cb6f46d54a0130f678c6097240a2bd |
| 1739 | Author: Daniel Drake <dsd@gentoo.org> |
| 1740 | Date: Sat Jan 5 21:59:40 2008 +0000 |
| 1741 | |
| 1742 | Remove more fpusb remnants |
| 1743 | |
| 1744 | commit dcd5e0075fa84601d997d29bb60bc8370b97a375 |
| 1745 | Author: Daniel Drake <dsd@gentoo.org> |
| 1746 | Date: Fri Jan 4 01:19:07 2008 +0000 |
| 1747 | |
| 1748 | TODO update |
| 1749 | |
| 1750 | Start listing the things I want to discuss/reconsider before 1.0 API |
| 1751 | is final. |
| 1752 | |
| 1753 | commit de53d972ed9c1f21f86ffd585560520e986f212c |
| 1754 | Author: Daniel Drake <dsd@gentoo.org> |
| 1755 | Date: Fri Jan 4 01:17:06 2008 +0000 |
| 1756 | |
| 1757 | Rename some _devh_ functions |
| 1758 | |
| 1759 | Decided against naming functions in this style based on advice from |
| 1760 | Tim Roberts. Still not really sure about devh naming in general. |
| 1761 | |
| 1762 | commit 132d84bf8ae34bd0892b3d8207a53d87cc0a9cba |
| 1763 | Author: Daniel Drake <dsd@gentoo.org> |
| 1764 | Date: Fri Jan 4 01:13:25 2008 +0000 |
| 1765 | |
| 1766 | Use "transfer" instead of "msg" |
| 1767 | |
| 1768 | Based on a discussion with Tim Roberts. |
| 1769 | |
| 1770 | commit ea6af58a8643137eb01eeb7d2f65404e390feb1b |
| 1771 | Author: Daniel Drake <dsd@gentoo.org> |
| 1772 | Date: Fri Jan 4 00:57:00 2008 +0000 |
| 1773 | |
| 1774 | Plural of bus is busses |
| 1775 | |
| 1776 | SF #1633043 |
| 1777 | |
| 1778 | commit dbb3fd871e3972b4e670f3161e7cd2f58f357600 |
| 1779 | Author: Daniel Drake <dsd@gentoo.org> |
| 1780 | Date: Fri Jan 4 00:54:57 2008 +0000 |
| 1781 | |
| 1782 | API renaming: remove fpi and usb_ stuff |
| 1783 | |
| 1784 | fpi changed to usbi. |
| 1785 | |
| 1786 | We should not expose structures with prefix "usb_" in the public |
| 1787 | namespace as it is quite likely there will be some conflict somewhere. |
| 1788 | Instead, using "libusb_" should be safer. |
| 1789 | |
| 1790 | commit e9364d72151ae1de9cce4175f330fe1529f02511 |
| 1791 | Author: Daniel Drake <dsd@gentoo.org> |
| 1792 | Date: Fri Jan 4 00:40:49 2008 +0000 |
| 1793 | |
| 1794 | Rename to libusb-1.0 |
| 1795 | |
| 1796 | I've taken over the libusb project, and what was previously known as fpusb |
| 1797 | will eventually be released as libusb-1.0. |
| 1798 | |
| 1799 | commit f4ed0b6a3e0a20555fdfc9e9b8be7e8a2c8ae192 |
| 1800 | Author: Daniel Drake <dsd@gentoo.org> |
| 1801 | Date: Sun Dec 30 22:09:50 2007 +0000 |
| 1802 | |
| 1803 | Compiler flags update |
| 1804 | |
| 1805 | Also fix some warnings |
| 1806 | |
| 1807 | commit d3d671efe27491baf1d2ceadbfeeb4e7428d6f9a |
| 1808 | Author: Daniel Drake <dsd@gentoo.org> |
| 1809 | Date: Sun Dec 30 22:05:17 2007 +0000 |
| 1810 | |
| 1811 | Fix signalfd structure size on x86 |
| 1812 | |
| 1813 | ptr is a 64-bit value |
| 1814 | |
| 1815 | commit dfac6756faad76071a92945e3e3a16b8a841a7ea |
| 1816 | Author: Daniel Drake <dsd@gentoo.org> |
| 1817 | Date: Wed Dec 12 23:32:54 2007 +0000 |
| 1818 | |
| 1819 | Don't pass ctrl_msg/bulk_msg structures through async paths |
| 1820 | |
| 1821 | These are easiest to construct on the stack, but in the async case it is |
| 1822 | likely that the submitting function returns (hence stack is destroyed) |
| 1823 | before URB completion. |
| 1824 | |
| 1825 | commit a8d2881eb7c273892acf2ff6e1f3f552631d1d11 |
| 1826 | Author: Daniel Drake <dsd@gentoo.org> |
| 1827 | Date: Mon Dec 3 23:29:22 2007 +0000 |
| 1828 | |
| 1829 | Add functionality to access device descriptors |
| 1830 | |
| 1831 | commit b5a7a411087b37a6bf56b4688a9e3290daaed153 |
| 1832 | Author: Daniel Drake <dsd@gentoo.org> |
| 1833 | Date: Mon Dec 3 22:27:21 2007 +0000 |
| 1834 | |
| 1835 | Rename fpusb_device_handle functions to devh |
| 1836 | |
| 1837 | Prevents confusion with functions that work on fpusb_dev structures. |
| 1838 | |
| 1839 | commit eacf4505a473df0f521d1b69d5b45448f0709e2d |
| 1840 | Author: Daniel Drake <dsd@gentoo.org> |
| 1841 | Date: Sun Dec 2 23:16:57 2007 +0000 |
| 1842 | |
| 1843 | Add API to access poll fd |
| 1844 | |
| 1845 | Almost forgot... :) |
| 1846 | |
| 1847 | commit 852bba4754ec57679c823f33e8feba6e4a564cbe |
| 1848 | Author: Daniel Drake <dsd@gentoo.org> |
| 1849 | Date: Wed Nov 28 13:48:45 2007 +0000 |
| 1850 | |
| 1851 | Initial commit |
| 1852 | |
| 1853 | Basic library structure which supports enumerating detected USB devices |