kmx git

Commit	Date	Message
e7fdadfc	2012-02-15T14:52:29	Automatically choose phatk kernel for bitalign non-gcn ATI cards, and then only select poclbm if SDK2.6 is detected.
6a785946	2012-02-15T14:47:02	Make SDK 2.6 warning and advice big and bold.
23c01bc7	2012-02-13T13:19:04	Make output buffer write only as per Diapolo's suggestion.
b2b5083b	2012-02-13T12:22:35	Microoptimise phatk kernel on return code.
fd05341a	2012-02-13T10:39:26	Do not loop indefinitely setting poclbm kernel to load a binary.
d689cfbd	2012-02-13T10:06:26	Try to load a binary if we've defaulted to the poclbm kernel on SDK2.6
3057b701	2012-02-13T09:59:29	Use the poclbm kernel on SDK2.6 with bitalign devices only if there is no binary available.
2c33f122	2012-02-13T08:34:44	Whitelist ATI SDK 2.6 to use the poclbm kernel by default.
fb99c8d5	2012-02-12T21:38:45	The longstanding generation of a zero sized binary appears to be due to the OpenCL library putting the binary in a RANDOM SLOT amongst 4 possible binary locations. Iterate over each of them after building from source till the real binary is found and use that.
56907db2	2012-02-12T18:21:30	Fix harmless warnings with -Wsign-compare to allow cgminer to build with -W.
405a2120	2012-02-11T20:11:18	Remove unnecessary check for opt_debug on every invocation of applog at LOG_DEBUG and place the check in applog().
60c70145	2012-02-11T16:41:41	Retain cl program after successfully loading a binary image.
55bd031d	2012-02-11T16:38:55	Variable unused after this so remove setting it.
1c1b8bec	2012-02-11T15:58:07	BFI INT patching is not necessarily true on binary loading of files and not true on ATI SDK2.6+. Report bitalign instead.
f2d5db0c	2012-02-10T16:45:35	Use only working kernels by default.
59d3d011	2012-02-10T14:33:40	Implement diablo kernel support and try to make it work.
95a989da	2012-02-10T13:18:16	Conflicting entries of cl_kernel may have been causing problems, and automatically chosen kernel type was not being passed on. Rename the enum to cl_kernels and store the chosen kernel in each clState.
e6cf96ad	2012-02-10T10:28:45	ALlow much longer filenames for kernels to load properly.
4822cca7	2012-02-10T10:23:06	Allow different kernels to be used by different devices and fix the logic fail of overcorrecting on last commit with !strstr.
196e8a0f	2012-02-10T09:10:57	Fix kernel selection process and build error.
47a09cea	2012-02-09T15:15:03	added OpenCL >= 1.1 detection code, in preparation of OpenCL 1.1 global offset parameter support
cb7145b1	2012-02-08T13:45:56	Add basic build ability with diakgcn and put all kernel names in configure.ac to avoid changing them in mutliple places.
6776b0ea	2012-02-10T16:45:35	Use only working kernels by default.
2270b4e0	2012-02-10T14:33:40	Implement diablo kernel support and try to make it work.
02c94272	2012-02-10T13:18:16	Conflicting entries of cl_kernel may have been causing problems, and automatically chosen kernel type was not being passed on. Rename the enum to cl_kernels and store the chosen kernel in each clState.
35ea31b1	2012-02-10T10:28:45	ALlow much longer filenames for kernels to load properly.
8af2365e	2012-02-10T10:23:06	Allow different kernels to be used by different devices and fix the logic fail of overcorrecting on last commit with !strstr.
2b23805e	2012-02-10T09:10:57	Fix kernel selection process and build error.
ed7210af	2012-02-09T15:15:03	added OpenCL >= 1.1 detection code, in preparation of OpenCL 1.1 global offset parameter support
a6c6866a	2012-02-08T13:45:56	Add basic build ability with diakgcn and put all kernel names in configure.ac to avoid changing them in mutliple places.
53c1e9ae	2012-02-04T15:15:57	Allow the OpenCL platform ID to be chosen with --gpu-platform.
a4f47812	2012-02-04T14:47:23	Iterate over all platforms displaying their information and number of devices when --ndevs is called.
ebaa2be1	2012-02-03T18:19:39	Update poclbm kernel for better performance on GCN and new SDKs with bitalign support when not BFI INT patching. Update phatk kernel to work properly for non BFI INT patched kernels, providing support for phatk to run on GCN and non-ATI cards.
82af288e	2012-01-29T22:57:29	Revert "Fix various harmless warnings." This reverts commit a4b67f030fc0c7e2b18e79114a441c1e1617d5f8.
a4b67f03	2012-01-29T21:06:17	Fix various harmless warnings.
b8f845b4	2012-01-29T16:43:38	Display information about the opencl platform with verbose enabled.
5d5584f8	2012-01-29T16:31:03	Explicitly check for nvidia in opencl platform strings as well.
a3d90f84	2012-01-29T11:01:17	Default to poclbm kernel on Tahiti (7970) since phatk does not work, even though performance is sub-standard so that at least it will mine successfully by default.
31f6e8c7	2012-01-28T17:06:28	Unset prog_built after it is patched because it needs rebuilding.
1e503549	2012-01-28T16:29:19	Retain cl program after every possible place we might build the program.
25caca90	2012-01-28T16:26:53	Revert "Don't explicitly retain the cl program as it is of no benefit to do so and may lead to problems when trying to release the program." This reverts commit 32910463a3124265b56aca48a6c0fbb107ccfb70. Turns out this does help.
32910463	2012-01-26T20:53:35	Don't explicitly retain the cl program as it is of no benefit to do so and may lead to problems when trying to release the program.
d18d5564	2012-01-26T20:39:35	Do not attempt to build the program that becomes the kernel twice. This could have been leading to failures on initialising cl.
c87460b3	2012-01-26T19:42:57	Typo.
2ecabd85	2012-01-26T19:38:15	Some opencl compilers have issues with no spaces after -D in the compiler options.
77e9b1c2	2012-01-26T13:06:39	Use calloced stack memory for CompilerOptions to ensure sprintf writes to the beginning of the char.
d7aac254	2012-01-26T11:44:42	Whitelist 79x0 cards to prefer no vectors as they perform better without.
3d4cfce8	2012-01-24T20:23:44	Instead of using the BFI_INT patching hack on any device reporting cl_amd_media_ops, create a whitelist of devices that need it. This should enable GCN architectures (ATI 79xx cards) to work properly.
6442c1ab	2012-01-22T20:36:57	Style police.
0719d407	2012-01-22T17:09:06	Clean up on failure to load a binary kernel.
fb0c580b	2011-10-15T13:29:44	Go to kernel build should we fail to clCreateProgramWithBinary instead of failing on that device. Should fix the windows problems with devices not initialising.
2053de6d	2011-09-06T10:11:34	Add the directory name from the arguments cgminer was called from as well to allow it running from a relative pathname.
5848c110	2011-08-29T00:16:58	Confusion over the variable name for number of devices was passing a bogus value which likely was causing the zero sized binary issue.
3567b69e	2011-08-26T10:20:02	Remove fragile source patching for bitalign, vectors et. al and simply pass it with the compiler options.
3d5f5554	2011-08-25T14:42:03	Allow a custom kernel path to be entered on the command line.
413d9709	2011-08-25T13:59:46	Make cgminer look in the install directory for the .cl files making make install work correctly.
48180b69	2011-08-25T13:10:53	Fail gracefully if unable to open the opencl files.
6d10ef2f	2011-08-22T10:17:23	Bump version numbers of kernels to indicate slightly different versions.
4beade37	2011-08-18T22:42:37	Retain the program immediately after it's created from source.
082e20df	2011-08-18T22:34:03	Explicitly tell the compiler to retain the program to minimise the chance of the zero sized binary errors.
0f782ba6	2011-08-17T15:47:18	Update poclbm kernel to FF sized mask and only check that range.
c40f51c7	2011-08-17T15:06:59	Move to cgminer style buffer return and file naming convention and fix a compiler warning.
d15d225a	2011-08-16T23:19:46	Changed phatk version to 2.2
42d49ffd	2011-08-15T20:28:25	Revert "Restart threads by abstracting out the clcontext initialisation and using that instead of probing all cards." This reverts commit 8f186e61e250e71bd606cabb52795eaa0c9ad423.
cf543507	2011-08-15T20:27:02	Revert "Preinitialise the devices only once on startup." This reverts commit 071a0ad2f156ab492ebea6c5a60a1e49a62466de.
b1289a01	2011-08-15T20:26:46	Revert "Move the non cl_ variables into the cgpu info struct to allow creating a new cl state on reinit, preserving known GPU variables." This reverts commit 28880d0dc7c601ee4479921502b66e913e38e36d.
28880d0d	2011-08-13T20:54:20	Move the non cl_ variables into the cgpu info struct to allow creating a new cl state on reinit, preserving known GPU variables. Create a new context from scratch in initCQ in case something was corrupted to maximise our chance of succesfully creating a new worker thread.
071a0ad2	2011-08-12T13:00:25	Preinitialise the devices only once on startup.
8f186e61	2011-07-30T16:59:54	Restart threads by abstracting out the clcontext initialisation and using that instead of probing all cards.
4365896b	2011-07-29T10:17:36	Release the command queue created after we've copied the binary data.
283d5d23	2011-07-29T10:09:24	Create a command queue from the program created from source which allows us to flush the command queue in the hope it will not generate a zero sized binary any more.
2e37e337	2011-07-24T10:58:03	Out of order command queue may fail on osx. Try without if it fails.
4cd12aa8	2011-07-24T09:04:56	Fix harmless warning.
a9e1a255	2011-07-23T15:15:46	Make it possible to select the choice of kernel on the command line.
116a9dc0	2011-07-23T14:17:25	Update phatk kernel to one with new parameters for slightly less overhead again. Make the queue kernel parameters call a function pointer to select phatk or poclbm.
1c67f606	2011-07-21T10:07:29	Sometimes the cl compiler generates zero sized binaries and only a reboot seems to fix it.
7b13812e	2011-07-21T09:58:28	Kernels are safely flushed in a way that allows out of order execution to work.
a7707a26	2011-07-18T10:42:24	Rename the poclbm file to ensure a new binary is built since.
eea05c05	2011-07-15T13:04:25	Update kernel with a shorter output path, and use 4k output buffer to match OS page sizes.
857902a1	2011-07-12T22:23:03	Commit a new phatk kernel renamed to force new binary building and add proper support in makefiles.
0c910673	2011-07-10T00:30:12	Set max preferred size to 256 to prevent lying cards from crashing when no worksize is set.
826cc480	2011-07-08T11:58:04	Opcode should be ULL.
efebee5a	2011-07-06T16:47:29	Fix the case where there are no GPUs, and exit if they give errors. If there are no GPUs, set nDevs to 0 not -1 (status is set to an unhelpful -1001 here on my laptop, so we can't rely on a particular status value). Also, if nDevs is -1, exit rather than screwing up later.
a636a674	2011-07-05T21:31:41	Merge branch 'cgminer' of git://github.com/ckolivas/cgminer into cgminer
52d6e7ca	2011-07-05T21:31:24	Fixed fread issues under Windows.
a93b22c6	2011-07-05T17:34:54	Make it possible to build without GPU mining by picking up HAVE_OPENCL from config.h.
5d301c8b	2011-07-02T10:22:09	Make a binary load failure build from source.
821da37c	2011-07-04T13:49:28	Add hardware name to binary kernel name allowing for unique kernels for different cards on the same machine.
13b43cfa	2011-07-03T00:28:51	Update copyright and authors.
594b38b8	2011-07-02T13:46:17	Fix redefinition of gnu source.
ec831917	2011-06-25T04:43:37	Build on windows using mingw32.
4d730577	2011-06-30T10:36:19	Build binaries with unique filenames from the kernel generated and save them. Try to load this cached binary if it matches on next kernel instantiation. This speeds up start-up dramatically, and has a unique kernel binary for different kernel configurations.
2b6e8416	2011-06-29T23:38:16	Use a buffer of up to 512 * 4 integers when retrieving work from the GPU. This allows each local thread id to have one slot to put any positive results into, thus making overlapping results far less likely. Thus races will be much rarer, allowing more threads. It should also pick up blocks close to each other more reliably and hopefully decrease the number of rejects and opencl errors. Do the search over the buffer entirely in a separate thread to allow the GPU to stay as busy as possible. Detach threads from themselves to prevent unlucky even where dereferencing occurs by freeing the data that stores the thread info.
60f0bb19	2011-06-30T15:47:17	Temporarily back out binary building till it's working more reliably.
973b2199	2011-06-30T08:58:07	Tidy.
3aa5be4f	2011-07-01T01:14:43	Reinstate binary kernel loading with fixes. Build binaries with unique filenames from the kernel generated and save them. Try to load this cached binary if it matches on next kernel instantiation. This speeds up start-up dramatically, and has a unique kernel binary for different kernel configurations.
a095f0fa	2011-06-30T14:30:10	Broke source generated program. Fix.
6374e0fa	2011-06-28T21:11:04	Import the phatk kernel. Enable it only for hardware with amd media ops for now since it crashes nvidia et. al. Fallback to the poclbm kernel for the rest. Try harder to avoid stale blocks around longpoll detecting new blocks.
2dbb3944	2011-06-27T22:05:03	Base was being set wrongly meaning we were repeating searches and the rate was actually lower than displayed :( Tweak Ma with new changes. Change default vectors to 2 since it's faster than 4 even when 4 is reported as preferred.
06f39506	2011-06-26T08:49:50	Fix typo which prevented BFI INT patch working on multi-GPUs.

e7fdadfc

2012-02-15T14:52:29

Automatically choose phatk kernel for bitalign non-gcn ATI cards, and then only select poclbm if SDK2.6 is detected.

6a785946

2012-02-15T14:47:02

Make SDK 2.6 warning and advice big and bold.

23c01bc7

2012-02-13T13:19:04

Make output buffer write only as per Diapolo's suggestion.

b2b5083b

2012-02-13T12:22:35

Microoptimise phatk kernel on return code.

fd05341a

2012-02-13T10:39:26

Do not loop indefinitely setting poclbm kernel to load a binary.

d689cfbd

2012-02-13T10:06:26

Try to load a binary if we've defaulted to the poclbm kernel on SDK2.6

3057b701

2012-02-13T09:59:29

Use the poclbm kernel on SDK2.6 with bitalign devices only if there is no binary available.

2c33f122

2012-02-13T08:34:44

Whitelist ATI SDK 2.6 to use the poclbm kernel by default.

fb99c8d5

2012-02-12T21:38:45

The longstanding generation of a zero sized binary appears to be due to the OpenCL library putting the binary in a RANDOM SLOT amongst 4 possible binary locations. Iterate over each of them after building from source till the real binary is found and use that.

56907db2

2012-02-12T18:21:30

Fix harmless warnings with -Wsign-compare to allow cgminer to build with -W.

405a2120

2012-02-11T20:11:18

Remove unnecessary check for opt_debug on every invocation of applog at LOG_DEBUG and place the check in applog().

60c70145

2012-02-11T16:41:41

Retain cl program after successfully loading a binary image.

55bd031d

2012-02-11T16:38:55

Variable unused after this so remove setting it.

1c1b8bec

2012-02-11T15:58:07

BFI INT patching is not necessarily true on binary loading of files and not true on ATI SDK2.6+. Report bitalign instead.

f2d5db0c

2012-02-10T16:45:35

Use only working kernels by default.

59d3d011

2012-02-10T14:33:40

Implement diablo kernel support and try to make it work.

95a989da

2012-02-10T13:18:16

Conflicting entries of cl_kernel may have been causing problems, and automatically chosen kernel type was not being passed on. Rename the enum to cl_kernels and store the chosen kernel in each clState.

e6cf96ad

2012-02-10T10:28:45

ALlow much longer filenames for kernels to load properly.

4822cca7

2012-02-10T10:23:06

Allow different kernels to be used by different devices and fix the logic fail of overcorrecting on last commit with !strstr.

196e8a0f

2012-02-10T09:10:57

Fix kernel selection process and build error.

47a09cea

2012-02-09T15:15:03

added OpenCL >= 1.1 detection code, in preparation of OpenCL 1.1 global offset parameter support

cb7145b1

2012-02-08T13:45:56

Add basic build ability with diakgcn and put all kernel names in configure.ac to avoid changing them in mutliple places.

6776b0ea

2012-02-10T16:45:35

Use only working kernels by default.

2270b4e0

2012-02-10T14:33:40

Implement diablo kernel support and try to make it work.

02c94272

2012-02-10T13:18:16

Conflicting entries of cl_kernel may have been causing problems, and automatically chosen kernel type was not being passed on. Rename the enum to cl_kernels and store the chosen kernel in each clState.

35ea31b1

2012-02-10T10:28:45

ALlow much longer filenames for kernels to load properly.

8af2365e

2012-02-10T10:23:06

Allow different kernels to be used by different devices and fix the logic fail of overcorrecting on last commit with !strstr.

2b23805e

2012-02-10T09:10:57

Fix kernel selection process and build error.

ed7210af

2012-02-09T15:15:03

added OpenCL >= 1.1 detection code, in preparation of OpenCL 1.1 global offset parameter support

a6c6866a

2012-02-08T13:45:56

Add basic build ability with diakgcn and put all kernel names in configure.ac to avoid changing them in mutliple places.

53c1e9ae

2012-02-04T15:15:57

Allow the OpenCL platform ID to be chosen with --gpu-platform.

a4f47812

2012-02-04T14:47:23

Iterate over all platforms displaying their information and number of devices when --ndevs is called.

ebaa2be1

2012-02-03T18:19:39

Update poclbm kernel for better performance on GCN and new SDKs with bitalign support when not BFI INT patching. Update phatk kernel to work properly for non BFI INT patched kernels, providing support for phatk to run on GCN and non-ATI cards.

82af288e

2012-01-29T22:57:29

Revert "Fix various harmless warnings." This reverts commit a4b67f030fc0c7e2b18e79114a441c1e1617d5f8.

a4b67f03

2012-01-29T21:06:17

Fix various harmless warnings.

b8f845b4

2012-01-29T16:43:38

Display information about the opencl platform with verbose enabled.

5d5584f8

2012-01-29T16:31:03

Explicitly check for nvidia in opencl platform strings as well.

a3d90f84

2012-01-29T11:01:17

Default to poclbm kernel on Tahiti (7970) since phatk does not work, even though performance is sub-standard so that at least it will mine successfully by default.

31f6e8c7

2012-01-28T17:06:28

Unset prog_built after it is patched because it needs rebuilding.

1e503549

2012-01-28T16:29:19

Retain cl program after every possible place we might build the program.

25caca90

2012-01-28T16:26:53

Revert "Don't explicitly retain the cl program as it is of no benefit to do so and may lead to problems when trying to release the program." This reverts commit 32910463a3124265b56aca48a6c0fbb107ccfb70. Turns out this does help.

32910463

2012-01-26T20:53:35

Don't explicitly retain the cl program as it is of no benefit to do so and may lead to problems when trying to release the program.

d18d5564

2012-01-26T20:39:35

Do not attempt to build the program that becomes the kernel twice. This could have been leading to failures on initialising cl.

c87460b3

2012-01-26T19:42:57

Typo.

2ecabd85

2012-01-26T19:38:15

Some opencl compilers have issues with no spaces after -D in the compiler options.

77e9b1c2

2012-01-26T13:06:39

Use calloced stack memory for CompilerOptions to ensure sprintf writes to the beginning of the char.

d7aac254

2012-01-26T11:44:42

Whitelist 79x0 cards to prefer no vectors as they perform better without.

3d4cfce8

2012-01-24T20:23:44

Instead of using the BFI_INT patching hack on any device reporting cl_amd_media_ops, create a whitelist of devices that need it. This should enable GCN architectures (ATI 79xx cards) to work properly.

6442c1ab

2012-01-22T20:36:57

Style police.

0719d407

2012-01-22T17:09:06

Clean up on failure to load a binary kernel.

fb0c580b

2011-10-15T13:29:44

Go to kernel build should we fail to clCreateProgramWithBinary instead of failing on that device. Should fix the windows problems with devices not initialising.

2053de6d

2011-09-06T10:11:34

Add the directory name from the arguments cgminer was called from as well to allow it running from a relative pathname.

5848c110

2011-08-29T00:16:58

Confusion over the variable name for number of devices was passing a bogus value which likely was causing the zero sized binary issue.

3567b69e

2011-08-26T10:20:02

Remove fragile source patching for bitalign, vectors et. al and simply pass it with the compiler options.

3d5f5554

2011-08-25T14:42:03

Allow a custom kernel path to be entered on the command line.

413d9709

2011-08-25T13:59:46

Make cgminer look in the install directory for the .cl files making make install work correctly.

48180b69

2011-08-25T13:10:53

Fail gracefully if unable to open the opencl files.

6d10ef2f

2011-08-22T10:17:23

Bump version numbers of kernels to indicate slightly different versions.

4beade37

2011-08-18T22:42:37

Retain the program immediately after it's created from source.

082e20df

2011-08-18T22:34:03

Explicitly tell the compiler to retain the program to minimise the chance of the zero sized binary errors.

0f782ba6

2011-08-17T15:47:18

Update poclbm kernel to FF sized mask and only check that range.

c40f51c7

2011-08-17T15:06:59

Move to cgminer style buffer return and file naming convention and fix a compiler warning.

d15d225a

2011-08-16T23:19:46

Changed phatk version to 2.2

42d49ffd

2011-08-15T20:28:25

Revert "Restart threads by abstracting out the clcontext initialisation and using that instead of probing all cards." This reverts commit 8f186e61e250e71bd606cabb52795eaa0c9ad423.

cf543507

2011-08-15T20:27:02

Revert "Preinitialise the devices only once on startup." This reverts commit 071a0ad2f156ab492ebea6c5a60a1e49a62466de.

b1289a01

2011-08-15T20:26:46

Revert "Move the non cl_ variables into the cgpu info struct to allow creating a new cl state on reinit, preserving known GPU variables." This reverts commit 28880d0dc7c601ee4479921502b66e913e38e36d.

28880d0d

2011-08-13T20:54:20

Move the non cl_ variables into the cgpu info struct to allow creating a new cl state on reinit, preserving known GPU variables. Create a new context from scratch in initCQ in case something was corrupted to maximise our chance of succesfully creating a new worker thread.

071a0ad2

2011-08-12T13:00:25

Preinitialise the devices only once on startup.

8f186e61

2011-07-30T16:59:54

Restart threads by abstracting out the clcontext initialisation and using that instead of probing all cards.

4365896b

2011-07-29T10:17:36

Release the command queue created after we've copied the binary data.

283d5d23

2011-07-29T10:09:24

Create a command queue from the program created from source which allows us to flush the command queue in the hope it will not generate a zero sized binary any more.

2e37e337

2011-07-24T10:58:03

Out of order command queue may fail on osx. Try without if it fails.

4cd12aa8

2011-07-24T09:04:56

Fix harmless warning.

a9e1a255

2011-07-23T15:15:46

Make it possible to select the choice of kernel on the command line.

116a9dc0

2011-07-23T14:17:25

Update phatk kernel to one with new parameters for slightly less overhead again. Make the queue kernel parameters call a function pointer to select phatk or poclbm.

1c67f606

2011-07-21T10:07:29

Sometimes the cl compiler generates zero sized binaries and only a reboot seems to fix it.

7b13812e

2011-07-21T09:58:28

Kernels are safely flushed in a way that allows out of order execution to work.

a7707a26

2011-07-18T10:42:24

Rename the poclbm file to ensure a new binary is built since.

eea05c05

2011-07-15T13:04:25

Update kernel with a shorter output path, and use 4k output buffer to match OS page sizes.

857902a1

2011-07-12T22:23:03

Commit a new phatk kernel renamed to force new binary building and add proper support in makefiles.

0c910673

2011-07-10T00:30:12

Set max preferred size to 256 to prevent lying cards from crashing when no worksize is set.

826cc480

2011-07-08T11:58:04

Opcode should be ULL.

efebee5a

2011-07-06T16:47:29

Fix the case where there are no GPUs, and exit if they give errors. If there are no GPUs, set nDevs to 0 not -1 (status is set to an unhelpful -1001 here on my laptop, so we can't rely on a particular status value). Also, if nDevs is -1, exit rather than screwing up later.

a636a674

2011-07-05T21:31:41

Merge branch 'cgminer' of git://github.com/ckolivas/cgminer into cgminer

52d6e7ca

2011-07-05T21:31:24

Fixed fread issues under Windows.

a93b22c6

2011-07-05T17:34:54

Make it possible to build without GPU mining by picking up HAVE_OPENCL from config.h.

5d301c8b

2011-07-02T10:22:09

Make a binary load failure build from source.

821da37c

2011-07-04T13:49:28

Add hardware name to binary kernel name allowing for unique kernels for different cards on the same machine.

13b43cfa

2011-07-03T00:28:51

Update copyright and authors.

594b38b8

2011-07-02T13:46:17

Fix redefinition of gnu source.

ec831917

2011-06-25T04:43:37

Build on windows using mingw32.

4d730577

2011-06-30T10:36:19

Build binaries with unique filenames from the kernel generated and save them. Try to load this cached binary if it matches on next kernel instantiation. This speeds up start-up dramatically, and has a unique kernel binary for different kernel configurations.

2b6e8416

2011-06-29T23:38:16

Use a buffer of up to 512 * 4 integers when retrieving work from the GPU. This allows each local thread id to have one slot to put any positive results into, thus making overlapping results far less likely. Thus races will be much rarer, allowing more threads. It should also pick up blocks close to each other more reliably and hopefully decrease the number of rejects and opencl errors. Do the search over the buffer entirely in a separate thread to allow the GPU to stay as busy as possible. Detach threads from themselves to prevent unlucky even where dereferencing occurs by freeing the data that stores the thread info.

60f0bb19

2011-06-30T15:47:17

Temporarily back out binary building till it's working more reliably.

973b2199

2011-06-30T08:58:07

Tidy.

3aa5be4f

2011-07-01T01:14:43

Reinstate binary kernel loading with fixes. Build binaries with unique filenames from the kernel generated and save them. Try to load this cached binary if it matches on next kernel instantiation. This speeds up start-up dramatically, and has a unique kernel binary for different kernel configurations.

a095f0fa

2011-06-30T14:30:10

Broke source generated program. Fix.

6374e0fa

2011-06-28T21:11:04

Import the phatk kernel. Enable it only for hardware with amd media ops for now since it crashes nvidia et. al. Fallback to the poclbm kernel for the rest. Try harder to avoid stale blocks around longpoll detecting new blocks.

2dbb3944

2011-06-27T22:05:03

Base was being set wrongly meaning we were repeating searches and the rate was actually lower than displayed :( Tweak Ma with new changes. Change default vectors to 2 since it's faster than 4 even when 4 is reported as preferred.

06f39506

2011-06-26T08:49:50

Fix typo which prevented BFI INT patch working on multi-GPUs.

thodg/cgminer/ocl.c

ocl.c

Log