e9c1cb6 | Paul Zimmermann | 06 December 2016, 10:04:20 UTC | fixed default L1 cache size (merged commit 4ec9a94 from master) | 06 December 2016, 10:04:20 UTC |
68c5e3b | Paul Zimmermann | 05 December 2016, 22:01:47 UTC | openmp fix | 05 December 2016, 22:01:47 UTC |
44e9628 | Paul Zimmermann | 02 November 2016, 12:14:47 UTC | Revert "[purge] reduced col-min-index by a factor 2" This reverts commit 05a2796d95e8685c96fd4a05b448424848da2fe0. (should have been applied to master, not to ropt_dev) | 02 November 2016, 12:14:47 UTC |
05a2796 | Paul Zimmermann | 02 November 2016, 09:15:08 UTC | [purge] reduced col-min-index by a factor 2 | 02 November 2016, 09:15:08 UTC |
cf55d6b | Shi Bai | 29 August 2016, 15:56:16 UTC | current version for ropt_dev | 29 August 2016, 15:56:16 UTC |
ef9c4cd | Paul Zimmermann | 22 August 2016, 07:37:11 UTC | fixed typo | 22 August 2016, 07:37:11 UTC |
9dabdae | Emmanuel Thomé | 20 August 2016, 21:18:26 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr/gitroot/cado-nfs/cado-nfs | 20 August 2016, 21:18:26 UTC |
26d3bf0 | Laurent Grémy | 18 August 2016, 13:21:47 UTC | Forget one thing selection for the galois action. | 18 August 2016, 13:21:47 UTC |
3717191 | Laurent Grémy | 17 August 2016, 14:04:33 UTC | New special-q selection for the galois action. | 17 August 2016, 14:04:33 UTC |
9473917 | Pierrick Gaudry | 16 August 2016, 13:21:24 UTC | Support sieving a special q < 2^I. | 16 August 2016, 13:21:24 UTC |
05307cd | Pierrick Gaudry | 16 August 2016, 13:16:26 UTC | I need longer lines in parameters. | 16 August 2016, 13:16:26 UTC |
3a5a9f5 | Emmanuel Thomé | 12 August 2016, 06:59:59 UTC | most of last commit got accidentally uncommitted... | 12 August 2016, 06:59:59 UTC |
799e7e2 | Emmanuel Thomé | 11 August 2016, 15:00:46 UTC | Make change of commit 9d78938 conditional. | 11 August 2016, 15:29:50 UTC |
186316f | Emmanuel Thomé | 11 August 2016, 14:54:00 UTC | helps corner cases | 11 August 2016, 14:54:00 UTC |
9d78938 | Emmanuel Thomé | 11 August 2016, 13:07:04 UTC | prevent random_matrix from creating test matrices with zero columns | 11 August 2016, 13:07:08 UTC |
1b03cbe | Pierrick Gaudry | 11 August 2016, 12:40:46 UTC | A bit faster like this. | 11 August 2016, 12:40:46 UTC |
8ea66d1 | Pierrick Gaudry | 11 August 2016, 12:40:25 UTC | New option to las to completely bypass the cofactorization and print survivors. | 11 August 2016, 12:40:25 UTC |
f28c4ae | Emmanuel Thomé | 10 August 2016, 12:30:37 UTC | early safety check: avoid zero columns in matrix | 10 August 2016, 12:30:37 UTC |
68158ec | Emmanuel Thomé | 10 August 2016, 12:07:05 UTC | some minor changes to bwc-ptrace.m | 10 August 2016, 12:07:05 UTC |
bf8b507 | Emmanuel Thomé | 10 August 2016, 09:46:30 UTC | indentation | 10 August 2016, 09:46:30 UTC |
9c50203 | Emmanuel Thomé | 10 August 2016, 09:40:31 UTC | indentation | 10 August 2016, 09:40:31 UTC |
ac82639 | Emmanuel Thomé | 10 August 2016, 09:18:12 UTC | mpfq update | 10 August 2016, 09:18:12 UTC |
c88120d | Emmanuel Thomé | 10 August 2016, 08:41:54 UTC | mpfq update | 10 August 2016, 08:41:54 UTC |
f7a4945 | Emmanuel Thomé | 09 August 2016, 17:04:29 UTC | improve one diagnostic message in bwc gather | 10 August 2016, 08:36:09 UTC |
b8a5e2f | Pierrick Gaudry | 08 August 2016, 11:51:02 UTC | MAXDEGREE 12 is not supported | 08 August 2016, 11:51:02 UTC |
485a4d3 | Pierrick Gaudry | 08 August 2016, 09:18:37 UTC | oops, forgot the tests | 08 August 2016, 09:18:37 UTC |
f3a8c08 | Pierrick Gaudry | 08 August 2016, 09:17:18 UTC | The nfs-hd subdir is not yet ready for continuous integration! | 08 August 2016, 09:17:18 UTC |
a802394 | Laurent Grémy | 08 August 2016, 06:38:53 UTC | Merge branch 'master' into hd | 08 August 2016, 06:38:53 UTC |
1464d7b | Laurent Grémy | 08 August 2016, 06:37:29 UTC | Use the full version of check_relations. | 08 August 2016, 06:37:29 UTC |
9ef2841 | Emmanuel Thomé | 01 August 2016, 11:44:54 UTC | more cpubinding examples | 01 August 2016, 11:56:36 UTC |
e7c5e8d | Emmanuel Thomé | 01 August 2016, 11:44:54 UTC | more cpubinding examples | 01 August 2016, 11:44:54 UTC |
df532d6 | Emmanuel Thomé | 31 July 2016, 09:09:12 UTC | remove more clutter | 31 July 2016, 09:09:12 UTC |
0686793 | Emmanuel Thomé | 31 July 2016, 08:47:06 UTC | fix wrongly designed feature about tasks.xxx.run=false | 31 July 2016, 08:48:03 UTC |
036dc82 | Shi Bai | 27 July 2016, 13:08:21 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/cado-nfs/cado-nfs | 27 July 2016, 13:08:21 UTC |
3775173 | Shi Bai | 27 July 2016, 13:08:18 UTC | fix bug computing exp_alpha | 27 July 2016, 13:08:18 UTC |
d9baf1c | Emmanuel Thomé | 27 July 2016, 08:55:03 UTC | Try to adopt a relatively uniform policy regarding TMPDIR TMPDIR should be the place where all mktemp-created stuff goes. And directories below that shouldn't be called TMPDIR in turn (unless we wish that to be exported as the parent of subsequently created directories). I also wish we could generalize the construct: wdir=$(mktemp -d ${TMPDIR-/tmp}/cado.XXXXXXXX) cleanup() { if ! [ "$CADO_DEBUG" ] ; then rm -rf $wdir ; fi ; } trap cleanup EXIT with the open question of whether we should wipe out in case of error (which the above does, unless CADO_DEBUG is set), or explicitly leave stuff in place in that case (as done, e.g. in tests/sieve/strategies/test_main_script.sh) | 27 July 2016, 08:58:55 UTC |
53e7dab | Paul Zimmermann | 26 July 2016, 15:33:13 UTC | added option -I (alternative to -area) to polyselect_ropt | 26 July 2016, 15:33:13 UTC |
654ef80 | Paul Zimmermann | 26 July 2016, 08:13:10 UTC | avoid errors in polyselect_ropt with tiny ropt_effort | 26 July 2016, 08:13:10 UTC |
2791120 | Shi Bai | 25 July 2016, 17:48:32 UTC | fix a bug found by PZ | 25 July 2016, 17:48:32 UTC |
c5bb690 | Shi Bai | 20 July 2016, 23:32:47 UTC | fix doc | 20 July 2016, 23:32:47 UTC |
9378860 | Shi Bai | 20 July 2016, 23:11:30 UTC | add another experimental feature for tuning, not used for now | 20 July 2016, 23:11:30 UTC |
a38560e | Shi Bai | 20 July 2016, 15:20:23 UTC | combine linear and quad functions to one | 20 July 2016, 15:20:23 UTC |
f4acb79 | Laurent Grémy | 19 July 2016, 17:56:02 UTC | Merge branch 'master' into hd | 19 July 2016, 17:56:02 UTC |
1ab8e11 | Laurent Grémy | 19 July 2016, 17:55:19 UTC | Test primality during check_relations. | 19 July 2016, 17:55:19 UTC |
25754a3 | Shi Bai | 19 July 2016, 08:43:31 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/cado-nfs/cado-nfs | 19 July 2016, 08:43:31 UTC |
9a80794 | Paul Zimmermann | 19 July 2016, 07:02:56 UTC | in dlpolyselect, only print record scores so far | 19 July 2016, 07:02:56 UTC |
90ced02 | Paul Zimmermann | 19 July 2016, 06:50:48 UTC | now in 2016 | 19 July 2016, 06:50:48 UTC |
b4d5a16 | Laurent Grémy | 13 July 2016, 12:16:05 UTC | Merge branch 'master' into hd | 13 July 2016, 12:16:05 UTC |
20b9ba2 | Laurent Grémy | 13 July 2016, 12:11:45 UTC | MAYBE_UNUSED for variables that allow to do stats for the space sieve. | 13 July 2016, 12:11:45 UTC |
743dbd4 | Shi Bai | 10 July 2016, 16:03:04 UTC | another strategy for tuning lognorm bound | 10 July 2016, 16:03:04 UTC |
2b79e55 | Shi Bai | 08 July 2016, 12:08:54 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/cado-nfs/cado-nfs | 08 July 2016, 12:08:54 UTC |
f501ed8 | Emmanuel Thomé | 07 July 2016, 10:50:20 UTC | fix wrong constructs (mere warnings, but still) | 07 July 2016, 10:50:43 UTC |
6da0f2c | Paul Zimmermann | 07 July 2016, 09:34:40 UTC | removed (now) unused functions | 07 July 2016, 09:34:40 UTC |
58d448e | Pierrick Gaudry | 07 July 2016, 09:04:50 UTC | Remember that facul statistics are not thread-safe. | 07 July 2016, 09:04:50 UTC |
5a60a3b | Pierrick Gaudry | 07 July 2016, 09:04:30 UTC | Make clang happy. | 07 July 2016, 09:04:30 UTC |
f25c681 | Paul Zimmermann | 06 July 2016, 11:57:34 UTC | document --bindir | 06 July 2016, 11:57:34 UTC |
1c0a97c | Emmanuel Thomé | 05 July 2016, 10:45:28 UTC | update mpi.cmake to accomodate subtle variations of config query tools. mpicc -showme:version sometimes prints on stderr ompi_info has changed (-v versus -V ... ; only the bare call works uniformly...). | 05 July 2016, 10:59:26 UTC |
7da8d30 | Emmanuel Thomé | 05 July 2016, 10:45:28 UTC | update mpi.cmake | 05 July 2016, 10:58:43 UTC |
608598f | Shi Bai | 04 July 2016, 15:41:00 UTC | adjust param | 04 July 2016, 15:41:00 UTC |
4f4fdc4 | Emmanuel Thomé | 04 July 2016, 11:24:23 UTC | binary compatibility issue | 04 July 2016, 11:24:23 UTC |
883b6ae | Shi Bai | 04 July 2016, 08:33:08 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/cado-nfs/cado-nfs | 04 July 2016, 08:33:08 UTC |
70c9c56 | Shi Bai | 04 July 2016, 08:24:17 UTC | more tuning on quadratic | 04 July 2016, 08:24:17 UTC |
4696991 | Emmanuel Thomé | 04 July 2016, 08:21:46 UTC | extra protection | 04 July 2016, 08:21:46 UTC |
b7cee36 | Shi Bai | 04 July 2016, 08:00:44 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/cado-nfs/cado-nfs | 04 July 2016, 08:00:44 UTC |
b124e37 | Emmanuel Thomé | 04 July 2016, 07:58:44 UTC | some c++ template joy. Ugly. | 04 July 2016, 07:58:44 UTC |
74b980f | Shi Bai | 03 July 2016, 16:03:01 UTC | pre-tune for quadratic | 03 July 2016, 16:03:01 UTC |
b7f1b58 | Shi Bai | 03 July 2016, 15:53:51 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/cado-nfs/cado-nfs | 03 July 2016, 15:53:51 UTC |
ba6ee83 | Emmanuel Thomé | 03 July 2016, 15:41:59 UTC | better for static builds, apparently | 03 July 2016, 15:41:59 UTC |
4147356 | Shi Bai | 03 July 2016, 15:32:59 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr//gitroot/cado-nfs/cado-nfs | 03 July 2016, 15:32:59 UTC |
e55b039 | Emmanuel Thomé | 03 July 2016, 15:28:39 UTC | Merge branch 'better-matmul-modp' | 03 July 2016, 15:28:39 UTC |
8c6e73f | Shi Bai | 03 July 2016, 14:01:16 UTC | also tuning incr on lognorm | 03 July 2016, 14:01:16 UTC |
17d18ae | Shi Bai | 02 July 2016, 15:57:07 UTC | more changes | 02 July 2016, 15:57:07 UTC |
c9f1433 | Shi Bai | 01 July 2016, 14:48:32 UTC | more changes | 01 July 2016, 14:48:32 UTC |
3dfead1 | Laurent Grémy | 30 June 2016, 17:52:37 UTC | New skew LLL implementation. | 30 June 2016, 17:52:37 UTC |
353deb8 | Shi Bai | 30 June 2016, 16:04:30 UTC | more changes | 30 June 2016, 16:04:30 UTC |
81e0c99 | Shi Bai | 30 June 2016, 15:50:08 UTC | recursive method for linear tuning | 30 June 2016, 15:50:08 UTC |
b86b061 | Emmanuel Thomé | 29 June 2016, 16:43:06 UTC | got trapped by C struct padding :-(( | 29 June 2016, 16:43:06 UTC |
242b7a1 | Emmanuel Thomé | 29 June 2016, 14:39:18 UTC | support spmv_test for non-square matrices | 29 June 2016, 14:39:18 UTC |
3f90409 | Paul Zimmermann | 29 June 2016, 10:12:54 UTC | use median to get better estimations | 29 June 2016, 10:12:54 UTC |
18b5aeb | Pierrick Gaudry | 28 June 2016, 16:11:53 UTC | Another try... | 28 June 2016, 16:11:53 UTC |
52a71c3 | Pierrick Gaudry | 28 June 2016, 15:52:20 UTC | Revert "Fix bug when using bkthresh1 and we encounter a very skewed q." This reverts commit ce3cf10edc7d1e6a952e5d8f9b926d0c3c30b929. | 28 June 2016, 15:52:20 UTC |
ce3cf10 | Pierrick Gaudry | 28 June 2016, 15:35:03 UTC | Fix bug when using bkthresh1 and we encounter a very skewed q. The next q was perturbated by it. | 28 June 2016, 15:35:03 UTC |
8c8cd06 | Paul Zimmermann | 28 June 2016, 11:25:18 UTC | consider samples of fixed size | 28 June 2016, 11:25:18 UTC |
e52b18b | Pierrick Gaudry | 28 June 2016, 09:46:58 UTC | ...and allow the python script to see it. | 28 June 2016, 09:46:58 UTC |
f21873a | Pierrick Gaudry | 28 June 2016, 09:43:51 UTC | This maxfull bug drives me crazy. Fixing it properly will take some time. A new parameter, so that the user can survive in the meantime. | 28 June 2016, 09:43:51 UTC |
91c282e | Paul Zimmermann | 28 June 2016, 09:28:01 UTC | average (beta,eta) values to get less variability | 28 June 2016, 09:28:01 UTC |
66e31de | Paul Zimmermann | 28 June 2016, 05:52:16 UTC | consider full samples only in polyselect (Weibull estimation) | 28 June 2016, 05:52:16 UTC |
ae9c7bf | Paul Zimmermann | 27 June 2016, 14:22:21 UTC | mis-placed clear | 27 June 2016, 14:22:21 UTC |
c349b6e | Pierrick Gaudry | 27 June 2016, 11:43:58 UTC | was way too optimistic... | 27 June 2016, 11:43:58 UTC |
54bb763 | Paul Zimmermann | 27 June 2016, 10:58:52 UTC | fix error in previous commit | 27 June 2016, 10:58:52 UTC |
6416e2d | Paul Zimmermann | 27 June 2016, 10:12:55 UTC | new Weibull estimation via extreme values | 27 June 2016, 10:12:55 UTC |
6fce023 | Pierrick Gaudry | 27 June 2016, 08:52:15 UTC | Yet another attempt to deal with the maxfull problem. | 27 June 2016, 08:52:15 UTC |
1fcfd1d | Pierrick Gaudry | 27 June 2016, 08:22:09 UTC | Print small sieving info only if verbose. | 27 June 2016, 08:51:29 UTC |
1741490 | Pierrick Gaudry | 27 June 2016, 08:21:35 UTC | Back to 50 samples in mat size estimates. | 27 June 2016, 08:51:29 UTC |
9263d4a | Pierrick Gaudry | 27 June 2016, 08:21:14 UTC | Make -skipped param visible from the python. | 27 June 2016, 08:51:29 UTC |
b8f4e75 | Paul Zimmermann | 27 June 2016, 07:27:35 UTC | Merge branch 'master' of git+ssh://scm.gforge.inria.fr/git/cado-nfs/cado-nfs Conflicts: README parameters/factor/params.c110 parameters/factor/params.c190 | 27 June 2016, 07:27:35 UTC |
e49963c | Paul Zimmermann | 27 June 2016, 07:23:14 UTC | better parameters | 27 June 2016, 07:23:14 UTC |
7d3b51a | Paul Zimmermann | 27 June 2016, 07:22:36 UTC | reduce nq to 256 | 27 June 2016, 07:22:36 UTC |
52ac6b9 | Paul Zimmermann | 27 June 2016, 07:21:54 UTC | mention support for hwloc | 27 June 2016, 07:21:54 UTC |
c63002b | Pierrick Gaudry | 27 June 2016, 07:09:28 UTC | Bugfix in duplicate suppression: mfb was not properly taken into account. | 27 June 2016, 07:09:28 UTC |