d1a8317 | edwards | 14 July 2009, 18:41:00 UTC | Induced changes | 14 July 2009, 18:41:00 UTC |
a8be58a | edwards | 14 July 2009, 18:39:00 UTC | Info on 3.35.0 | 14 July 2009, 18:39:00 UTC |
eb7d3a6 | edwards | 14 July 2009, 18:34:48 UTC | Changed to 3.35.0 | 14 July 2009, 18:34:48 UTC |
3c3eb01 | edwards | 13 July 2009, 04:54:08 UTC | Periodic update. SKIPPED: ChangeLog | 13 July 2009, 04:54:08 UTC |
3f52651 | jbulava | 13 July 2009, 03:59:49 UTC | More work on this. Not even close to compiling yet. | 13 July 2009, 03:59:49 UTC |
ccd2a1e | edwards | 12 July 2009, 00:56:14 UTC | Periodic update. SKIPPED: ChangeLog | 12 July 2009, 00:56:14 UTC |
0245115 | jbulava | 12 July 2009, 00:45:36 UTC | Bug fix. | 12 July 2009, 00:45:36 UTC |
987e475 | jbulava | 12 July 2009, 00:44:43 UTC | STOCH_LAPH_QUARK is ready to go. More work on STOCH_LAPH_BARYON to come. | 12 July 2009, 00:44:43 UTC |
6c16939 | edwards | 09 July 2009, 04:55:30 UTC | Periodic update. SKIPPED: ChangeLog | 09 July 2009, 04:55:30 UTC |
b0c553e | jbulava | 09 July 2009, 02:13:17 UTC | Changed the name of the diluted prop routine (per Colin's request). Also added a VERY rough start to a routine for make baryon source and sink functions from the diluted propagators. | 09 July 2009, 02:13:17 UTC |
c37751e | edwards | 09 July 2009, 00:53:40 UTC | Periodic update. SKIPPED: ChangeLog | 09 July 2009, 00:53:40 UTC |
899ac04 | jbulava | 08 July 2009, 21:44:23 UTC | bug fix. | 08 July 2009, 21:44:23 UTC |
486f1a2 | jbulava | 08 July 2009, 21:06:54 UTC | Some more revisions. | 08 July 2009, 21:06:54 UTC |
419a1a9 | edwards | 08 July 2009, 20:55:04 UTC | Periodic update. SKIPPED: ChangeLog | 08 July 2009, 20:55:04 UTC |
a4b1c02 | bjoo | 08 July 2009, 18:52:36 UTC | Induced changes | 08 July 2009, 18:52:36 UTC |
bdf612a | bjoo | 08 July 2009, 18:50:57 UTC | Bumped version to 3-34-1 | 08 July 2009, 18:50:57 UTC |
7ca6960 | bjoo | 08 July 2009, 18:46:47 UTC | Minor tweaks to the info echoed by solvers. Made BiCGStab and IBiCGStab output timings. Added regressions | 08 July 2009, 18:46:47 UTC |
af7f2ff | bjoo | 08 July 2009, 18:45:52 UTC | Added Regressions for IBICGStab and Reliable IBICGSTAB. Reduced volumes for some other leapfrog tests to speed up how much wallclock time they use | 08 July 2009, 18:45:52 UTC |
fff6c06 | edwards | 08 July 2009, 00:54:10 UTC | Periodic update. SKIPPED: ChangeLog | 08 July 2009, 00:54:10 UTC |
b1b8bbf | bjoo | 07 July 2009, 21:00:14 UTC | OK. Minor bugfixes. set stream_ps back to store_ps. Fixed multithreading issues and added regression tests for new linop solvers | 07 July 2009, 21:00:14 UTC |
2b0f1b7 | edwards | 07 July 2009, 20:55:08 UTC | Periodic update. SKIPPED: ChangeLog | 07 July 2009, 20:55:08 UTC |
44f17b7 | bjoo | 07 July 2009, 19:13:20 UTC | Made reliable_ibicgstab more stable by reinitializing sigma as well. Debugged the kernels so that they can now run in multihreaded mode too | 07 July 2009, 19:13:20 UTC |
f79e4d0 | edwards | 06 July 2009, 20:55:03 UTC | Periodic update. SKIPPED: ChangeLog | 06 July 2009, 20:55:03 UTC |
f363c4a | bjoo | 06 July 2009, 19:02:57 UTC | Added Reliable IBiCGSTab | 06 July 2009, 19:02:57 UTC |
eede824 | bjoo | 06 July 2009, 19:02:34 UTC | First attempt at reliable IBCGStab. Seems to get the right answer with threads=1. Need to debug kernels for QMT_NUM_THREADS more than 1 | 06 July 2009, 19:02:34 UTC |
e1ec13f | edwards | 03 July 2009, 20:54:55 UTC | Periodic update. SKIPPED: ChangeLog | 03 July 2009, 20:54:55 UTC |
07c513c | jbulava | 03 July 2009, 16:51:02 UTC | More checks.... | 03 July 2009, 16:51:02 UTC |
bef2948 | edwards | 03 July 2009, 00:53:37 UTC | Periodic update. SKIPPED: ChangeLog | 03 July 2009, 00:53:37 UTC |
85205da | bjoo | 02 July 2009, 22:11:01 UTC | Added BiCRStab. Not really a win it seems | 02 July 2009, 22:11:01 UTC |
ab36526 | edwards | 02 July 2009, 20:54:11 UTC | Periodic update. SKIPPED: ChangeLog | 02 July 2009, 20:54:11 UTC |
e22946d | bjoo | 02 July 2009, 18:24:48 UTC | Moved IBiCGStab to its own solver (IBICGSTAB_INVERTER) and put back old BiCGStab so both are available. Checked kernels in double prec, looks good. Changed Inline Hadron Aggregate etc so that the InlineLaplaceEig stuff doesnt compile unless BUILD_LAPACK is defined with an --enable-lapack=lapack | 02 July 2009, 18:24:48 UTC |
e703898 | edwards | 02 July 2009, 16:53:26 UTC | Periodic update. SKIPPED: ChangeLog | 02 July 2009, 16:53:26 UTC |
d4c9d8c | bjoo | 02 July 2009, 15:09:20 UTC | Seems to now work in scalar. On my laptop has multithread issues. Check on 7n & 9q for parallel operation | 02 July 2009, 15:09:20 UTC |
bdc2aad | edwards | 02 July 2009, 12:53:55 UTC | Periodic update. SKIPPED: ChangeLog | 02 July 2009, 12:53:55 UTC |
c5fc10c | edwards | 02 July 2009, 11:10:03 UTC | Changed to using a qdp-lapack routine. This fixes the double build. | 02 July 2009, 11:10:03 UTC |
a99e228 | edwards | 01 July 2009, 20:54:48 UTC | Periodic update. SKIPPED: ChangeLog | 01 July 2009, 20:54:48 UTC |
f7ed925 | bjoo | 01 July 2009, 20:36:10 UTC | OK! IBigstab actually works now. Found all bugs in the paper and PETSc. Now have one kernel - zvupdates in double to debug still. Then to recheck single again | 01 July 2009, 20:36:10 UTC |
16ca4f3 | edwards | 01 July 2009, 16:54:18 UTC | Periodic update. SKIPPED: ChangeLog | 01 July 2009, 16:54:18 UTC |
a1df87a | jbulava | 01 July 2009, 14:10:01 UTC | Added more tests. Seems to work. Still without implicit restart. | 01 July 2009, 14:10:01 UTC |
810b45e | edwards | 01 July 2009, 00:53:38 UTC | Periodic update. SKIPPED: ChangeLog | 01 July 2009, 00:53:38 UTC |
2f57379 | jbulava | 30 June 2009, 21:10:13 UTC | Compiles and runs. Need to implement restarting now. | 30 June 2009, 21:10:13 UTC |
4bc284f | edwards | 30 June 2009, 20:55:26 UTC | Periodic update. SKIPPED: ChangeLog | 30 June 2009, 20:55:26 UTC |
dc63706 | jbulava | 30 June 2009, 20:24:58 UTC | Compiles, but gives runtime error. | 30 June 2009, 20:24:58 UTC |
920c31e | bjoo | 30 June 2009, 19:48:10 UTC | Restored BiCGStab | 30 June 2009, 19:48:10 UTC |
966cc76 | edwards | 30 June 2009, 16:54:23 UTC | Periodic update. SKIPPED: ChangeLog | 30 June 2009, 16:54:23 UTC |
e713eea | bjoo | 30 June 2009, 15:57:08 UTC | Induced changes from IBiCGStab SKIPPED: ChangeLog | 30 June 2009, 15:57:08 UTC |
c03d635 | bjoo | 30 June 2009, 15:52:10 UTC | Switched BiCGStab over to IBICGStab. Needs thorough testing, but seems OK in single and double on my laptop. Something weird is going on with multithreads in scalar Will check in parscalar elsewhere | 30 June 2009, 15:52:10 UTC |
f245d51 | jbulava | 30 June 2009, 15:51:50 UTC | First crack at a routine to do all inversions needed in the diluted LapH scheme. Both sources and sinks are inserted into a single DB file. Still have to smear the sink before writing. Also, at the moment the source is stored as a full lattice fermion. This is extremely wasteful if only a single t0 is to be used, but won't matter if we end up doing all t0's. | 30 June 2009, 15:51:50 UTC |
54f39a8 | jbulava | 30 June 2009, 15:48:10 UTC | Support for keys of LapH-diluted props. | 30 June 2009, 15:48:10 UTC |
ca4efc9 | edwards | 29 June 2009, 16:55:34 UTC | Periodic update. SKIPPED: ChangeLog | 29 June 2009, 16:55:34 UTC |
28f8a69 | jbulava | 29 June 2009, 14:35:26 UTC | Currently compiles, solves tridiagonal evp, but more work to be done. | 29 June 2009, 14:35:26 UTC |
5c6bb07 | edwards | 28 June 2009, 20:52:59 UTC | Periodic update. SKIPPED: ChangeLog | 28 June 2009, 20:52:59 UTC |
1fc7212 | edwards | 28 June 2009, 17:24:36 UTC | New regression output. | 28 June 2009, 17:24:36 UTC |
0f8fcc2 | edwards | 28 June 2009, 17:23:49 UTC | Fixed paths. | 28 June 2009, 17:23:49 UTC |
1cc36c2 | edwards | 25 June 2009, 16:54:34 UTC | Periodic update. SKIPPED: ChangeLog | 25 June 2009, 16:54:34 UTC |
5ed644c | bjoo | 25 June 2009, 13:03:54 UTC | Added timing info to chrono predictor | 25 June 2009, 13:03:54 UTC |
4fedfc9 | edwards | 24 June 2009, 20:53:06 UTC | Periodic update. SKIPPED: ChangeLog | 24 June 2009, 20:53:06 UTC |
3be15d3 | jbulava | 24 June 2009, 20:01:52 UTC | Induced changes | 24 June 2009, 20:01:52 UTC |
5bd3830 | jbulava | 24 June 2009, 19:59:33 UTC | Diluted source in which noise is entered in the LapH subspace only. Seems to compile. | 24 June 2009, 19:59:33 UTC |
513722d | jbulava | 24 June 2009, 19:55:40 UTC | *** empty log message *** | 24 June 2009, 19:55:40 UTC |
84b4cfa | edwards | 23 June 2009, 20:55:08 UTC | Periodic update. SKIPPED: ChangeLog | 23 June 2009, 20:55:08 UTC |
0e55099 | edwards | 23 June 2009, 20:34:00 UTC | Updating input/output. | 23 June 2009, 20:34:00 UTC |
a797e65 | jbulava | 23 June 2009, 19:18:59 UTC | Krylov vectors are orthogonal after big fix. | 23 June 2009, 19:18:59 UTC |
88485fa | jbulava | 23 June 2009, 19:01:26 UTC | Fixed bug. | 23 June 2009, 19:01:26 UTC |
f3f42d8 | edwards | 23 June 2009, 16:54:57 UTC | Periodic update. SKIPPED: ChangeLog | 23 June 2009, 16:54:57 UTC |
4c8fe36 | jbulava | 23 June 2009, 16:47:22 UTC | Induced changes. | 23 June 2009, 16:47:22 UTC |
8d84736 | jbulava | 23 June 2009, 15:22:48 UTC | Added a simple test of orthogonality | 23 June 2009, 15:22:48 UTC |
00d80de | jbulava | 23 June 2009, 15:12:40 UTC | Added first vesrion of the lanczos routine. Seems to compile, no restart yet. | 23 June 2009, 15:12:40 UTC |
14dd592 | edwards | 17 June 2009, 20:54:35 UTC | Periodic update. SKIPPED: ChangeLog | 17 June 2009, 20:54:35 UTC |
bf3cc9e | jbulava | 17 June 2009, 18:11:35 UTC | More work on this. Not ready to compile yet. | 17 June 2009, 18:11:35 UTC |
5508e9e | edwards | 16 June 2009, 00:54:52 UTC | Periodic update. SKIPPED: ChangeLog | 16 June 2009, 00:54:52 UTC |
066243e | jbulava | 15 June 2009, 18:04:29 UTC | More work. | 15 June 2009, 18:04:29 UTC |
87835bb | jbulava | 15 June 2009, 15:20:12 UTC | First crack at lanczos algorithm to obtain laplacian eigenpairs | 15 June 2009, 15:20:12 UTC |
063c2b5 | bjoo | 12 June 2009, 19:53:41 UTC | Added another buncha kernels | 12 June 2009, 19:53:41 UTC |
b29c312 | bjoo | 12 June 2009, 19:52:44 UTC | Added another kernel | 12 June 2009, 19:52:44 UTC |
2da2119 | edwards | 11 June 2009, 16:59:53 UTC | Periodic update. SKIPPED: ChangeLog | 11 June 2009, 16:59:53 UTC |
0baa5b3 | bjoo | 11 June 2009, 15:20:54 UTC | Solvers seem OK. Do not use ssed-clover for now. It seems to produce nonreproducibility in solvers | 11 June 2009, 15:20:54 UTC |
1232ecf | edwards | 10 June 2009, 21:03:09 UTC | Periodic update. SKIPPED: ChangeLog | 10 June 2009, 21:03:09 UTC |
ac78bf7 | bjoo | 10 June 2009, 20:05:04 UTC | Added SSE Kernels for BICGSTAB. | 10 June 2009, 20:05:04 UTC |
2b85b82 | bjoo | 10 June 2009, 19:58:58 UTC | Commit before these disappear | 10 June 2009, 19:58:58 UTC |
5f615da | edwards | 05 June 2009, 16:59:56 UTC | Periodic update. SKIPPED: ChangeLog | 05 June 2009, 16:59:56 UTC |
9e77740 | bjoo | 05 June 2009, 13:08:29 UTC | Reorganized the Reliable BiCGStab so no mallocs in the main loop | 05 June 2009, 13:08:29 UTC |
b31737b | edwards | 04 June 2009, 21:00:49 UTC | Periodic update. SKIPPED: ChangeLog | 04 June 2009, 21:00:49 UTC |
1294bbc | bjoo | 04 June 2009, 20:33:20 UTC | Induced Changes | 04 June 2009, 20:33:20 UTC |
d7ffe65 | bjoo | 04 June 2009, 20:32:48 UTC | Changed so that the first solve takes psi as the initial guess when using chronology | 04 June 2009, 20:32:48 UTC |
b3a2937 | bjoo | 04 June 2009, 20:30:27 UTC | Made ratio_conv_conv take a NULL predictor as its default. It sets an initial guess first | 04 June 2009, 20:30:27 UTC |
c4134f5 | bjoo | 04 June 2009, 20:29:13 UTC | Added Null Predictor | 04 June 2009, 20:29:13 UTC |
11c7e46 | edwards | 03 June 2009, 21:00:28 UTC | Periodic update. SKIPPED: ChangeLog | 03 June 2009, 21:00:28 UTC |
8b66d67 | bjoo | 03 June 2009, 19:20:29 UTC | Added prototype bicgstab-kernels, to Clover Rel BiCGstab only. When these work out they can propagate to normal BiCGStab too. | 03 June 2009, 19:20:29 UTC |
268f269 | edwards | 02 June 2009, 17:00:10 UTC | Periodic update. SKIPPED: ChangeLog | 02 June 2009, 17:00:10 UTC |
232ffb6 | bjoo | 02 June 2009, 15:56:39 UTC | Changed chrono solver interface in 4D monomials. ChronoPredictors are now passed into the SysSolverMdagM objects through their new operator\(\). No changes to 5D SysSolverMdagMs or 5D Monomials. No Changes to Polyprec Monomials. Regressions appear to pass | 02 June 2009, 15:56:39 UTC |
ab13228 | edwards | 01 June 2009, 21:00:21 UTC | Periodic update. SKIPPED: ChangeLog | 01 June 2009, 21:00:21 UTC |
2a31f02 | bjoo | 01 June 2009, 20:39:36 UTC | Updated 4D Chrono predictors to handle 2 step processes. Monomial and MdagM updates next | 01 June 2009, 20:39:36 UTC |
dcd7ff4 | bjoo | 01 June 2009, 19:46:36 UTC | Initial chronology changes to support two step chronology | 01 June 2009, 19:46:36 UTC |
934ff6a | edwards | 01 June 2009, 17:00:22 UTC | Periodic update. SKIPPED: ChangeLog | 01 June 2009, 17:00:22 UTC |
27324fa | bjoo | 01 June 2009, 16:24:54 UTC | Reliable CG can now do chronology. Also Can turn off force monitors | 01 June 2009, 16:24:54 UTC |
5367255 | edwards | 28 May 2009, 17:00:19 UTC | Periodic update. SKIPPED: ChangeLog | 28 May 2009, 17:00:19 UTC |
009f73b | bjoo | 28 May 2009, 15:36:30 UTC | Slight changes in the reporting of the solvers | 28 May 2009, 15:36:30 UTC |
6a2939f | edwards | 22 May 2009, 20:59:51 UTC | Periodic update. SKIPPED: ChangeLog | 22 May 2009, 20:59:51 UTC |
427c8bb | bjoo | 22 May 2009, 19:50:38 UTC | Oopsies | 22 May 2009, 19:50:38 UTC |