|
|
08/24/07 05:09:03 |
@236
|
[236]
|
sbahra |
These close tickets #9 and #10. Patch submitted by Kun Xi. |
|
|
07/16/07 21:21:51 |
@235
|
[235]
|
sbahra |
It appears Fahad made a whopping mistake. :-) |
|
|
07/09/07 14:05:30 |
@234
|
[234]
|
sbahra |
Remove whitespace. |
|
|
07/09/07 14:00:41 |
@233
|
[233]
|
sbahra |
Some missing headers. Thanks to Brian for submitting (ticket #8). |
|
|
07/02/07 15:15:46 |
@232
|
[232]
|
sbahra |
Remove old comment. |
|
|
07/02/07 15:14:28 |
@231
|
[231]
|
sbahra |
Add upc_mem function definitions as it appears someone … |
|
|
06/18/07 19:44:21 |
@230
|
[230]
|
sbahra |
Remove sh_error, gcc-upc has issues with redefinition
of shared symbols. |
|
|
05/25/07 10:12:24 |
@228
|
[228]
|
sbahra |
Add gcc-upc and update copyright. |
|
|
05/25/07 10:08:41 |
@227
|
[227]
|
sbahra |
Copyright changes to include current year. |
|
|
05/25/07 04:42:07 |
@226
|
[226]
|
sbahra |
Don't forget to initialize AIOBUILD. |
|
|
05/25/07 04:40:13 |
@225
|
[225]
|
sbahra |
Typo fix. |
|
|
05/25/07 04:01:24 |
@224
|
[224]
|
sbahra |
Add empty flag to make sure blocksize of 0 can … |
|
|
05/25/07 02:43:17 |
@223
|
[223]
|
sbahra |
No real reason to link this. Break out library … |
|
|
05/25/07 01:21:28 |
@222
|
[222]
|
sbahra |
Make sure to use gmake rather than make. |
|
|
04/28/07 09:47:42 |
@221
|
[221]
|
sbahra |
Some whitespace clean-up. |
|
|
04/28/07 09:45:14 |
@220
|
[220]
|
sbahra |
Clarifying his comment round 2. |
|
|
04/28/07 09:43:48 |
@219
|
[219]
|
sbahra |
Clarify Kun's comment. |
|
|
04/28/07 09:39:34 |
@218
|
[218]
|
sbahra |
Remove this old note. |
|
|
04/28/07 09:37:33 |
@217
|
[217]
|
sbahra |
Make sure to make use of aiocb only if GIO_RT_AIO
is noted. |
|
|
04/11/07 01:46:10 |
@216
|
[216]
|
sbahra |
Fix a typo in memvec_esize configure option portion. |
|
|
03/25/07 00:35:21 |
@215
|
[215]
|
sbahra |
Include NFS work-around. |
|
|
03/24/07 22:06:29 |
@214
|
[214]
|
sbahra |
Indentation fixes, rid of redundant actions. |
|
|
03/24/07 22:04:33 |
@213
|
[213]
|
sbahra |
Offset hacks for local operations. Rid of some redundancy. |
|
|
03/24/07 19:10:40 |
@212
|
[212]
|
sbahra |
- Increment offset for local … |
|
|
03/24/07 18:46:17 |
@211
|
[211]
|
sbahra |
Increment offset for no GIO RT AIO. |
|
|
03/24/07 18:31:12 |
@210
|
[210]
|
sbahra |
Increment offset for all cases with no GIO_RT_AIO. |
|
|
03/22/07 17:45:05 |
@209
|
[209]
|
sbahra |
Revert back aio->number changes only include offset
calculation changes. |
|
|
03/22/07 06:26:19 |
@208
|
[208]
|
kunxi |
Remove the check for vsize, vsize=0 is legal. |
|
|
03/22/07 05:06:33 |
@207
|
[207]
|
kunxi |
Fix the setupvec bug, the calculation is not correct. |
|
|
03/21/07 20:01:00 |
@206
|
[206]
|
kunxi |
Make sure that list is not operated for common file
pointers. … |
|
|
03/21/07 17:49:39 |
@205
|
[205]
|
kunxi |
Remove the check of aio->number == 0 check, … |
|
|
03/20/07 00:17:36 |
@204
|
[204]
|
sbahra |
Remove O_DIRECT for now. Add it later as a hint. |
|
|
03/20/07 00:00:39 |
@203
|
[203]
|
sbahra |
Make use of pwrite and pread if a platform
does support it. |
|
|
03/18/07 23:29:30 |
@202
|
[202]
|
sbahra |
Make some changes based on Kun Xi's input. |
|
|
03/18/07 22:50:25 |
@201
|
[201]
|
sbahra |
Fix note :-) |
|
|
03/18/07 22:49:30 |
@200
|
[200]
|
sbahra |
Include a note about libaio |
|
|
03/18/07 22:40:13 |
@199
|
[199]
|
sbahra |
Add README file. |
|
|
03/18/07 22:04:06 |
@198
|
[198]
|
sbahra |
Modifications to include Tarek El-Ghazawi. |
|
|
03/18/07 21:14:01 |
@197
|
[197]
|
sbahra |
- Set flags in ftest for early exit … |
|
|
03/18/07 20:03:35 |
@196
|
[196]
|
sbahra |
Only link to -laio on Linux if AIO support is requested. |
|
|
03/17/07 22:13:30 |
@195
|
[195]
|
sbahra |
Remove comment with regards to O_SYNC. This does … |
|
|
03/17/07 21:38:02 |
@194
|
[194]
|
sbahra |
Check for WRONLY, on some platforms this crashes
if left to lio list I/O. |
|
|
03/17/07 18:24:14 |
@193
|
[193]
|
sbahra |
- Propagate marshal-related issues back to aio subsystem. … |
|
|
03/17/07 18:03:56 |
@192
|
[192]
|
sbahra |
- Set error on marshal failures. |
|
|
03/17/07 18:01:30 |
@191
|
[191]
|
kunxi |
Use generic list_aread for async list operation. |
|
|
03/17/07 17:52:16 |
@190
|
[190]
|
sbahra |
Make use of UPC_RANDOM and UPC_MARSHALED. |
|
|
03/17/07 17:02:12 |
@189
|
[189]
|
sbahra |
Don't require 1:1 mappings. |
|
|
03/17/07 16:46:54 |
@188
|
[188]
|
sbahra |
- White-space improvements
- Get rid of redundant code … |
|
|
03/17/07 16:32:06 |
@187
|
[187]
|
sbahra |
Minor refactoring. |
|
|
03/17/07 14:40:03 |
@186
|
[186]
|
sbahra |
Ignore potential errors in hints. As defined in
the specification. |
|
|
03/16/07 18:33:22 |
@185
|
[185]
|
sbahra |
Fix more of these ancient memory leaks left … |
|
|
03/16/07 18:21:06 |
@184
|
[184]
|
sbahra |
Fix Kun Xi's ancient memory leaks. |
|
|
03/16/07 15:11:51 |
@183
|
[183]
|
sbahra |
Include UPC_IO per-required by specification. |
|
|
03/16/07 10:52:13 |
@182
|
[182]
|
sbahra |
So, upc_lock_free is *not* collective. No wonder. … |
|
|
03/16/07 02:49:56 |
@181
|
[181]
|
sbahra |
Add UPC_TRUNC to fopen. |
|
|
03/16/07 02:25:33 |
@180
|
[180]
|
sbahra |
Add WRONLY check to list read. Add AIO backlog … |
|
|
03/16/07 01:46:16 |
@179
|
[179]
|
sbahra |
Linux :-) Thanks for not being coherent. |
|
|
03/16/07 01:18:43 |
@178
|
[178]
|
sbahra |
… |
|
|
03/16/07 01:11:09 |
@177
|
[177]
|
sbahra |
Add restrict to build. |
|
|
03/16/07 00:55:58 |
@176
|
[176]
|
sbahra |
Pass local descriptor. |
|
|
03/16/07 00:50:05 |
@175
|
[175]
|
sbahra |
Drop the barriers, try another approach. |
|
|
03/16/07 00:43:32 |
@174
|
[174]
|
sbahra |
Potential fix to a race. |
|
|
03/16/07 00:35:13 |
@173
|
[173]
|
sbahra |
Try a potential fix… |
|
|
03/16/07 00:23:53 |
@172
|
[172]
|
sbahra |
Handle enlargement better. |
|
|
03/16/07 00:13:46 |
@171
|
[171]
|
sbahra |
Handle errors in a way that platforms that do … |
|
|
03/15/07 23:21:21 |
@170
|
[170]
|
sbahra |
Whitespace clean-up. |
|
|
03/15/07 23:07:47 |
@169
|
[169]
|
sbahra |
Add restrict for some platforms. |
|
|
03/15/07 03:27:27 |
@168
|
[168]
|
sbahra |
There is no need for statuses to be strict. |
|
|
03/15/07 03:10:58 |
@167
|
[167]
|
sbahra |
Create required paths for installation. |
|
|
03/15/07 01:41:40 |
@166
|
[166]
|
sbahra |
"==" is not exactly portable. |
|
|
03/15/07 01:22:51 |
@165
|
[165]
|
sbahra |
Install the upcio-config tool in the appropriate
directory. |
|
|
03/15/07 00:51:52 |
@164
|
[164]
|
sbahra |
Check for error condition. |
|
|
03/14/07 08:11:59 |
@163
|
[163]
|
sbahra |
Use aio_return, to get the real read counts. |
|
|
03/14/07 04:11:35 |
@162
|
[162]
|
sbahra |
Add Abdullah. |
|
|
03/14/07 04:04:24 |
@161
|
[161]
|
sbahra |
Add upcio-config, to ease the process of compiling
upcio applications. |
|
|
03/14/07 01:58:52 |
@160
|
[160]
|
sbahra |
Use aio_return for real byte counts. Already done in fwait a long … |
|
|
03/14/07 01:36:32 |
@159
|
[159]
|
sbahra |
Handle tags more intelligently, re-use tags buffer. |
|
|
03/14/07 01:29:28 |
@158
|
[158]
|
sbahra |
Do not set tags to NULL on failure. |
|
|
03/14/07 01:27:40 |
@157
|
[157]
|
sbahra |
Tags is now local. |
|
|
03/14/07 00:35:53 |
@156
|
[156]
|
sbahra |
Set a default mode if UPC_CREATE is passed. umask … |
|
|
03/14/07 00:21:50 |
@155
|
[155]
|
sbahra |
Set tags outside. |
|
|
03/14/07 00:16:46 |
@154
|
[154]
|
sbahra |
Idiotic Solaris debugging has left some debug
statements in here :-) |
|
|
03/14/07 00:07:57 |
@153
|
[153]
|
sbahra |
Typo. |
|
|
03/13/07 23:57:34 |
@152
|
[152]
|
sbahra |
Revert back verbatim to original. Set tags to NULL … |
|
|
03/13/07 23:40:37 |
@151
|
[151]
|
sbahra |
Return back hint code. |
|
|
03/13/07 23:39:42 |
@150
|
[150]
|
sbahra |
Revert back to changing tags, older revision
of this code. |
|
|
03/13/07 20:01:21 |
@149
|
[149]
|
sbahra |
Definitely a bad way to approach this, switch to using … |
|
|
03/13/07 03:27:28 |
@148
|
[148]
|
kunxi |
Fix the mysterious create and other hints-oriented bugs. |
|
|
03/13/07 02:57:39 |
@147
|
[147]
|
kunxi |
Refine the upc_all_fwrite_list_shared with individual file pointer to get … |
|
|
03/13/07 02:53:16 |
@146
|
[146]
|
kunxi |
Fix the upc_all_fwrite_shared_list_async in individual file pointer with … |
|
|
03/10/07 03:52:45 |
@145
|
[145]
|
kunxi |
Fix the gupc_map_vec bug, polish the list shared read/write function call. |
|
|
03/05/07 03:33:40 |
@144
|
[144]
|
kunxi |
Fix the bug with elemsize enabled in upc_all_fwrite_shared_list in … |
|
|
03/04/07 07:08:30 |
@143
|
[143]
|
kunxi |
Fix the upc_all_fread_list_shared_async bug with elemsize support. |
|
|
03/04/07 05:21:59 |
@142
|
[142]
|
kunxi |
Fix typo. |
|
|
03/04/07 04:53:52 |
@141
|
[141]
|
kunxi |
Fix typo |
|
|
03/04/07 04:46:52 |
@140
|
[140]
|
kunxi |
Refactor the gupc_map_vec to make it more readable. |
|
|
03/04/07 04:30:44 |
@139
|
[139]
|
kunxi |
Fix the upc_all_fread_list_shared bug when elmesize is enabled. |
|
|
03/01/07 11:22:50 |
@138
|
[138]
|
sbahra |
These tools allow users to replace their bupc … |
|
|
02/23/07 03:39:56 |
@137
|
[137]
|
sbahra |
This will contain the bupc UPC I/O hiding tool as well … |
|
|
02/23/07 03:36:06 |
@136
|
[136]
|
sbahra |
This set of changes allows the elemsize option for memvec … |
