Skip to main content

Browse the archive

https://github.com/E3SM-Project/E3SM
06 April 2024, 03:10:57 UTC
sort by:
Revision Author Date Message Commit Date
909dede Junqi Yin 11 August 2017, 20:02:13 UTC Fix the way to add fflag for Melvin 11 August 2017, 20:02:13 UTC
cc1c133 Junqi Yin 11 August 2017, 17:32:33 UTC change optimization level to O0 and force initialization of local variables for Melvin machine. 11 August 2017, 17:32:33 UTC
eb7bfa2 noel 11 August 2017, 15:45:10 UTC Merge branch 'ndk/machinefiles/intel18-option-nersc' into next (PR #1710) Update modules for Cori after maintenance AND add intel18 option for Cori and Edison. After Cori's Aug 2017 maintenance, some modules needed to be updated. Module versions of edison, cori-haswell, and cori-knl are now more consistent. Added compiler option to all 3 NERSC machines to try the Intel 18 version. 11 August 2017, 15:45:10 UTC
07c1b1c noel 11 August 2017, 15:13:19 UTC After Cori maintenance in August 2017, some modules needed to be updated. Update craype, cray-mpich, pmi for cori-knl and cori-haswell. With cori-knl there are some apparent module bugs that I am working around to allow for the intel18 option. 11 August 2017, 15:13:19 UTC
8e11d46 Junqi Yin 10 August 2017, 20:53:55 UTC Revert "Merge branch 'wanggangsheng/lnd/alm-pflotran-bgc-rebase3' into next (PR #1649)" This reverts commit eae0ffd07cbbd39eddde17784460bb7ecfee516b, reversing changes made to 572b281e6f6148a3fa6ff58471740f2ae853a910. 10 August 2017, 20:53:55 UTC
8b87afb Michael Deakin 10 August 2017, 20:09:19 UTC Merge branch 'mt5555/cime/skybridge3' into next --npernode was set to the number of cores per node, independent of the number of threads used, resulting in oversubscribed nodes. Updated skybridge environment now rejects these jobs (and we shouldn't run this way anyway). Also: remove duplicate sbatch setting (so this is only set in one place) added support for Intel Broadwell machine 'ghost' removed support for decomissioned redsky [BFB] 10 August 2017, 20:09:19 UTC
7580b6c Mark Taylor 09 August 2017, 23:22:22 UTC replacing --npernode with mapping arguments copied from Anvil [BFB] 09 August 2017, 23:22:22 UTC
07c01d1 Mark Taylor 09 August 2017, 21:59:01 UTC bug fix: needed spaces {{ }} CIME syntax 09 August 2017, 21:59:01 UTC
f8b3ead Mark Taylor 09 August 2017, 19:07:30 UTC fix incorrect settings for --npernode on skybridge --npernode was set to the number of cores per node, independent of the number of threads used, resulting in oversubscribed nodes. Updated skybridge environment now rejects these jobs (and we shouldn't run this way anyway). Also: remove duplicate sbatch setting (so this is only set in one place) added support for Intel Broadwell machine 'ghost' removed support for decomissioned redsky [BFB] 09 August 2017, 19:07:30 UTC
e576bde Wuyin Lin 09 August 2017, 18:25:32 UTC Merge quantheory/atm/in_cloud_precip_gt_cld_frac (PR #1642) Add a limiter that requires the precipitation fraction calculated by MG2's 'in_cloud' method to be at least as great as the cloud fraction. As currently written, MG2 assumes that the fraction of a level containing cloud is a subset of the fraction of a level containing precipitation. Therefore we should require the precipitation fraction to be greater than or equal to the current cloud fraction for consistency. While the previous behavior should be considered a bug, the effect of this change is expected to be negligible for several reasons: The inconsistent state (in this case, precip_frac < cldm) is only possible if there is a significant cloud fraction in a level that has negligible cloud mass. Depending on the model configuration, this might be either rare or impossible. (In the latter case, this change is bit-for-bit.) Even if the precipitation fraction does drop below the cloud fraction, the two values will likely be similar. Processes that involve interactions between precipitation and clouds will be shut off if there is no cloud mass. The code controlling evaporation/sublimation code, while it does use both the precipitation and cloud fractions, contains a limiter that causes it to ignore the cloud fraction when the cloud mass is small. [non-BFB] 09 August 2017, 18:25:32 UTC
eae0ffd Junqi Yin 09 August 2017, 17:58:16 UTC Merge branch 'wanggangsheng/lnd/alm-pflotran-bgc-rebase3' into next (PR #1649) Migrate NGEE-Arctic CLM-PFLOTRAN to ACME Land Model through a generic clm_interface 1. clm_interface_dataType consists of biogeochemistry (bgcType) and thermal-hydrology (thType). 2. Rename namelist use_bgc_interface to use_clm_interface to include bgc & th interfaces. 3. Data passing between ALM & clm_interface is implemented in clm_interface_funcsMod.F90. 4. Data passing between clm_interface & PFLOTRAN is implemented in clm_interface_pflotranMod.F90. 5. Global-scale ALM-PFLOTRAN-bgc coupled run is tested. 6. The clm_interface is extended to incorporate PLFOTRAN thermal-hydrology in next step. 7. Add configuration for mach "cades" for alm-pflotran coupled run. 8. makefile: update "PETSC_LIB"-"variables" dir for petsc-git-version 1a9d3c3. [BFB] Conflicts: components/clm/src/biogeochem/CNNStateUpdate1Mod.F90 components/clm/src/biogeochem/CNNitrogenFluxType.F90 components/clm/src/biogeochem/PStateUpdate1Mod.F90 components/clm/src/biogeochem/PStateUpdate2Mod.F90 components/clm/src/biogeochem/PStateUpdate3Mod.F90 components/clm/src/biogeophys/HydrologyNoDrainageMod.F90 components/clm/src/main/clm_initializeMod.F90 components/clm/src/main/controlMod.F90 09 August 2017, 17:58:16 UTC
d985681 noel 09 August 2017, 17:38:57 UTC For all NERSC mmachines, add a `--compiler=intel18` option and continue making the machine settings more consistent. 09 August 2017, 17:38:57 UTC
ee8bef2 Gautam Bisht 09 August 2017, 15:41:41 UTC Merge branch 'jinyuntang/lnd/betr_v2int' (PR #1618) This replaces betr-v1 with the more modularized betr-v2. The code run successfully with gnu compiler on edison, however, it requires intel 'compiler 16.x to run and 15.x can only run in the debug model. This replacement does not affect other parts of ALM other than betr. [BFB] 09 August 2017, 15:41:41 UTC
572b281 Gautam Bisht 08 August 2017, 22:39:24 UTC Merge branch 'jinyuntang/lnd/betr_v2int' into next (PR #1618) This replaces betr-v1 with the more modularized betr-v2. The code run successfully with gnu compiler on edison, however, it requires intel 'compiler 16.x to run and 15.x can only run in the debug model. This replacement does not affect other parts of ALM other than betr. [BFB] 08 August 2017, 22:39:24 UTC
6d413c2 Wuyin Lin 08 August 2017, 20:32:41 UTC Enforce MG2 precondition for 'in_cloud' method Merge quantheory/atm/in_cloud_precip_gt_cld_frac (PR #1642) Add a limiter that requires the precipitation fraction calculated by MG2's 'in_cloud' method to be at least as great as the cloud fraction. As currently written, MG2 assumes that the fraction of a level containing cloud is a subset of the fraction of a level containing precipitation. Therefore we should require the precipitation fraction to be greater than or equal to the current cloud fraction for consistency. While the previous behavior should be considered a bug, the effect of this change is expected to be negligible for several reasons: The inconsistent state (in this case, precip_frac < cldm) is only possible if there is a significant cloud fraction in a level that has negligible cloud mass. Depending on the model configuration, this might be either rare or impossible. (In the latter case, this change is bit-for-bit.) Even if the precipitation fraction does drop below the cloud fraction, the two values will likely be similar. Processes that involve interactions between precipitation and clouds will be shut off if there is no cloud mass. The code controlling evaporation/sublimation code, while it does use both the precipitation and cloud fractions, contains a limiter that causes it to ignore the cloud fraction when the cloud mass is small. [non-BFB] 08 August 2017, 20:32:41 UTC
e07aaa6 Junqi Yin 08 August 2017, 20:19:16 UTC Merge branch 'bishtgautam/lnd/20tr-cn-test' into next (PR #1677) Adds a land test for 20th century transient A test with active land BGC for the 20th-century transient is added. This test will ensure the dynamic PFT capability in ALM will be protected. [BFB] 08 August 2017, 20:19:16 UTC
7a819b7 Junqi Yin 08 August 2017, 20:13:39 UTC Revert "Merge branch 'bishtgautam/lnd/20tr-cn-test' into next (PR #1677)" This reverts commit 1f5f07189ecc628847c1d7454852d56466b62530, reversing changes made to 35010ebd27b980bb8b999c14b86045ddf87bed0f. 08 August 2017, 20:13:39 UTC
1f5f071 Junqi Yin 08 August 2017, 19:56:59 UTC Merge branch 'bishtgautam/lnd/20tr-cn-test' into next (PR #1677) Adds a land test for 20th century transient A test with active land BGC for the 20th-century transient is added. This test will ensure the dynamic PFT capability in ALM will be protected. [BFB] 08 August 2017, 19:56:59 UTC
21ca187 Robert Jacob 08 August 2017, 19:44:10 UTC Merge branch 'brhillman/atm/add-ena-rrm' (PR #1269) Add configuration for a new regionally-refined (ne30 to ne120) grid over the Eastern North Atlantic. All components run on the RRM grid (enax4v1). All needed input files to run the new configuration have been uploaded to the input data repository. Supported compsets include only FC5AV1C-04P2 (72 levels), but RTM needs to be turned off (xmlchange RTM_MODE=NULL). Fixes ACME-Climate/ACME#1357 * brhillman/atm/add-ena-rrm: Set ATM_NCPL for ENA grid Set CAM_DYCORE to se for RRM grids Set CLM namelist definitions and defaults for ENA Set CAM namelist defaults for ENA grid Add ENA grid size to CAM configuration Add cime configuration for ENA grid 08 August 2017, 19:44:10 UTC
74bb8d9 Junqi Yin 08 August 2017, 18:51:41 UTC Merge branch 'bishtgautam/lnd/20tr-eca-compsets' (PR #1674) Adds 20th century transient I-Compsets for ECA Four 20th century transient compsets are added for the ECA model. Additionally, the land namelist build script is modified so that existing land initial conditions, which are valid for non-ECA case, aren't used when ECA is on. [NML] 08 August 2017, 18:51:41 UTC
f21ef4f Jinyun Tang 08 August 2017, 17:31:29 UTC bug fix for ibm and pgi compiler Now betr no longer reads in namelist from string, which seems to fix the failure from ibm and pgi comiplers. No changes to ALM are involved. 08 August 2017, 17:31:29 UTC
35010eb Robert Jacob 08 August 2017, 00:51:27 UTC Merge branch 'brhillman/atm/add-ena-rrm' into next (PR #1629) Add configuration for a new regionally-refined (ne30 to ne120) grid over the Eastern North Atlantic. All components run on the RRM grid (enax4v1). All needed input files to run the new configuration have been uploaded to the input data repository. Supported compsets include only FC5AV1C-04P2 (72 levels), but RTM needs to be turned off (xmlchange RTM_MODE=NULL). Fixes ACME-Climate/ACME#1357 * brhillman/atm/add-ena-rrm: Set ATM_NCPL for ENA grid Set CAM_DYCORE to se for RRM grids Set CLM namelist definitions and defaults for ENA Set CAM namelist defaults for ENA grid Add ENA grid size to CAM configuration Add cime configuration for ENA grid 08 August 2017, 00:51:27 UTC
5bc97a1 Azamat Mametjanov 07 August 2017, 23:30:42 UTC Merge branch 'worleyph/cime/add_NaNcheck_to_reprosum' into next (PR #1703) Check for NaNs and INFs in input to shr_reprosum_calc: In recent development NaNs and INFs are often first identified when passed to the shr_reprosum_calc routine, where they either lead to segmentation faults or to very slow performance (due to the interaction of NaNs and INFs with the reproducible sum logic). To more readily identify this error condition (and to prevent false attributions of error to shr_reprosum_calc), the input array is checked for the presence of NaNs and INFs. If found, an appropriate error message is output and the job is terminated. [BFB] 07 August 2017, 23:30:42 UTC
b537324 Patrick Worley 07 August 2017, 23:14:08 UTC Fix mistyping of variables New logical variables nan_check and inf_check were mistakenly declared as integer. 07 August 2017, 23:14:08 UTC
6725a21 Jon Wolfe 07 August 2017, 19:27:29 UTC Merge branch 'jonbob/mpas/update-070617' (PR #1630) Update MPAS components and framework, bring in threading optimizations This PR updates all MPAS submodules and framework, including: * bfb framework threading optimization from Abhinov MPAS#1237; * bfb mpas-o threading optimization from Abhinov MPAS#1235; * update path to new Albany install (cherry-picked from branch agsalin/machines/AlbanyNewInstall); * various framework and component updates. Tested with: * PET_Ln9.T62_oEC60to30v3.GMPAS-NYF.edison_intel * SMS.T62_oEC60to30v3.GMPAS-IAF.edison_intel (bfb with current master) * PET.T62_oEC60to30v3.GMPAS-NYF.titan Note: this commit is answer-changing only for compsets with MPASLI [non-BFB] 07 August 2017, 19:27:29 UTC
2936d42 Junqi Yin 07 August 2017, 17:20:26 UTC Merge branch 'bishtgautam/lnd/20tr-eca-compsets' into next (PR #1674) Adds 20th century transient I-Compsets for ECA Four 20th century transient compsets are added for the ECA model. Additionally, the land namelist build script is modified so that existing land initial conditions, which are valid for non-ECA case, aren't used when ECA is on. [NML] 07 August 2017, 17:20:26 UTC
f1f45ca Gautam Bisht 07 August 2017, 15:26:11 UTC Minor updates to the CLMbuildNamelist Adds back compatability when XML entry for initial condition did not include an identifier for nutrient competition pathway. All existing initial condition XML entries now include a new identifier (nu_com = 'RD'). If nu_com is undefined (=""), the model configuration is using SP (satellite phenology). In order to use existing initial condition nu_com = RD (instead of nu_comp='') is used to indentify the appropriate initial condition. 07 August 2017, 15:26:11 UTC
4ac2ee7 Patrick Worley 06 August 2017, 21:12:48 UTC Check for NaNs and INFs in input to shr_reprosum_calc In recent development NaNs and INFs are often first identified when passed to the shr_reprosum_calc routine, where they either lead to segmentation faults or to very slow performance (due to the interaction of NaNs and INFs with the reproducible sum logic). To more readily identify this error condition (and to prevent false attributions of error to shr_reprosum_calc), the input array is checked for the presence of NaNs and INFs. If found, an appropriate error message is output and the job is terminated. [BFB] *********1*********2*********3*********4*********5*********6*********7** Longer commit message body describing the commit. Can contain lists as follows: * Item 1 * Item 2 * Item 3 A good commit message should be written like an email, a subject followed by a blank line, followed by a more descriptive body. Can also contain a tag at the bottom describing what type of commit this is. [BFB] - Bit-For-Bit [FCC] - Flag Climate Changing [Non-BFB] - Non Bit-For-Bit [CC] - Climate Changing [NML] - Namelist Changing See confluence for a more detailed description about these tags. 06 August 2017, 21:12:48 UTC
0af35b6 Mark Taylor 04 August 2017, 19:22:33 UTC Merge mt5555/homme/skybridge to master (PR #1701) tweaks for standalone HOMME to compile on upgraded skybridge [BFB] 04 August 2017, 19:22:33 UTC
787884d Mark Taylor 04 August 2017, 19:14:29 UTC Merge branch 'mt5555/homme/skybridge' into next PR#1701 tweaks for standalone HOMME to compile on upgraded skybridge [BFB] 04 August 2017, 19:14:29 UTC
c9373a1 Mark Taylor 04 August 2017, 19:09:19 UTC tweaks for standalone HOMME to compile on upgraded skybridge [BFB] 04 August 2017, 19:09:19 UTC
0d1ee85 Junqi Yin 04 August 2017, 18:52:40 UTC Merge branch 'bishtgautam/lnd/20tr-eca-compsets' into next (PR #1674) Adds 20th century transient I-Compsets for ECA Four 20th century transient compsets are added for the ECA model. Additionally, the land namelist build script is modified so that existing land initial conditions, which are valid for non-ECA case, aren't used when ECA is on. [NML] 04 August 2017, 18:52:40 UTC
12e16e3 Jayesh Krishna 04 August 2017, 16:02:19 UTC Merge branch 'jayeshkrishna/blues_add_nag_pgi' (PR #1699) Adding support for pgi 15.7 and nag 6.0 on blues. [BFB] 04 August 2017, 16:02:19 UTC
9deae4b Jayesh Krishna 04 August 2017, 15:59:29 UTC Merge branch 'jayeshkrishna/blues_add_nag_pgi' into next (PR #1699) Adding support for pgi 15.7 and nag 6.0 on blues. [BFB] 04 August 2017, 15:59:29 UTC
e1c6cdc Jayesh Krishna 03 August 2017, 22:12:16 UTC Adding support for NAG compiler on blues Adding support for nag compiler (nag 6.0) on blues. * The -std=c99 flag is required because kissvec.c uses the "restrict" attribute that is part of C99 (The C compiler is gcc) * Only "mpich" is currently supported with the nag compiler on blues [BFB] 03 August 2017, 22:12:16 UTC
34c9909 noel 03 August 2017, 21:37:51 UTC Revert "Merge branch 'ndk/machinefiles/cori-knl-intelmpi-intel18b' into next (PR #1685)" This reverts commit d963c03d542a5d91915e7ee662a9e414cf81db46, reversing changes made to ef3387f8f2766cf106db77208cf6d58f52462002. 03 August 2017, 21:37:51 UTC
c1486ad Jon Wolfe 03 August 2017, 21:06:38 UTC Merge branch 'jonbob/mpas/update-070617' into next (PR #1630) Update MPAS components and framework, bring in threading optimizations This PR updates all MPAS submodules and framework, including: * bfb framework threading optimization from Abhinov MPAS#1237; * bfb mpas-o threading optimization from Abhinov MPAS#1235; * update path to new Albany install (cherry-picked from branch agsalin/machines/AlbanyNewInstall); * various framework and component updates. Tested with: * PET_Ln9.T62_oEC60to30v3.GMPAS-NYF.edison_intel * SMS.T62_oEC60to30v3.GMPAS-IAF.edison_intel (bfb with current master) * PET.T62_oEC60to30v3.GMPAS-NYF.titan Note: this commit is answer-changing only for compsets with MPASLI [non-BFB] 03 August 2017, 21:21:17 UTC
e43157e Jayesh Krishna 03 August 2017, 21:15:31 UTC Adding support for pgi compiler on blues Adding support for pgi compiler (pgi 15.7) on blues. [BFB] 03 August 2017, 21:15:31 UTC
b3d848e Jon Wolfe 03 August 2017, 21:06:01 UTC Revert "Revert "Merge branch 'jonbob/mpas/update-070617' into next (PR #1630)"" This reverts commit 23080db4aa0d92a00df1b021c075efc87b1c6036. Conflicts: cime/config/acme/machines/config_compilers.xml 03 August 2017, 21:06:01 UTC
02a7f60 Robert Jacob 03 August 2017, 16:28:01 UTC Merge branch 'azamat/mem-usage/swap-highwater' (PR #1639) Correct parsing of mem highwater values from cpl.log. Previously, memory resident set size was getting parsed while checking for memleaks. [BFB] Fixes ACME-Climate/ACME#1636 03 August 2017, 16:28:01 UTC
8af3502 Jinyun Tang 03 August 2017, 05:33:28 UTC change calling method of endrun in betr Now msg is defined before calling endrun in order to avoid building failure with pgi compiler. No changes of ALM are involved. 03 August 2017, 05:33:28 UTC
62d6582 Jinyun Tang 02 August 2017, 23:51:51 UTC change the namelist reader for betr Now namelist of betr is read per processor rather than per column. This is an attempt to fix the runtime failure with ibm compiler. No change to ALM is involved. 02 August 2017, 23:51:51 UTC
09daa64 Jayesh Krishna 01 August 2017, 22:12:17 UTC Re-organizing modules for blues Moving common modules (python, cmake) to a generic modules block 01 August 2017, 22:12:17 UTC
251cbc7 Jayesh Krishna 01 August 2017, 21:44:12 UTC Merge branch 'lcrc/machines/bebop' (PR #1662) Adding support for ANL Bebop cluster. (Re-merge of the branch was required because the previous merge missed a commit - 8aad11eea0e450eceb354ef509aa1b048e43668d ) [BFB] 01 August 2017, 21:44:12 UTC
0afc676 Jayesh Krishna 01 August 2017, 21:41:20 UTC Merge branch 'lcrc/machines/bebop' into next (PR #1662) Adding support for ANL Bebop cluster. (Re-merge was required because the previous merge missed a commit - 8aad11eea0e450eceb354ef509aa1b048e43668d ) [BFB] 01 August 2017, 21:41:20 UTC
7436dfd Jayesh Krishna 01 August 2017, 21:30:11 UTC Merge branch 'lcrc/machines/bebop' (PR #1662) Adding support for ANL Bebop cluster. [BFB] Conflicts: cime/config/acme/machines/config_batch.xml 01 August 2017, 21:30:11 UTC
0765934 Gautam Bisht 01 August 2017, 19:57:26 UTC Merge branch 'jinyuntang/lnd/betr_v2int' into next (PR #1618) This replaces betr-v1 with the more modularized betr-v2. The code run successfully with gnu compiler on edison, however, it requires intel 'compiler 16.x to run and 15.x can only run in the debug model. This replacement does not affect other parts of ALM other than betr. [BFB] 01 August 2017, 19:57:26 UTC
5396292 Jinyun Tang 01 August 2017, 19:52:43 UTC bug fix for skybridge failures Now ep_betr is always allocated to avoid runtime failure complaining ep_betr is not associated with a target. 01 August 2017, 19:52:43 UTC
a6d7316 James Foucar 01 August 2017, 19:14:47 UTC Merge branch 'jgfouca/cime/fix_st_archive_template' into master (PR #1692) Fix st_archive template. Not sure how we didn't catch this before. The st_archive was using an outdated way of handling the command line arguments. Fixes #1691 [BFB] * jgfouca/cime/fix_st_archive_template: Fix st_archive template. 01 August 2017, 19:15:25 UTC
cc55fbf James Foucar 01 August 2017, 19:14:47 UTC Merge branch 'jgfouca/cime/fix_st_archive_template' into next (PR #1692) Fix st_archive template. Not sure how we didn't catch this before. The st_archive was using an outdated way of handling the command line arguments. Fixes #1691 [BFB] * jgfouca/cime/fix_st_archive_template: Fix st_archive template. 01 August 2017, 19:14:47 UTC
bcff498 James Foucar 01 August 2017, 19:12:08 UTC Fix st_archive template. Not sure how we didn't catch this before. [BFB] 01 August 2017, 19:12:08 UTC
616802a Jon Wolfe 01 August 2017, 18:36:31 UTC Merge branch 'jonbob/cime/fix-cplalbav-setting' (PR #1652) Fix CPL_ALBAV setting for DATM/MPASO configurations. This PR fixes the setting for CPL_ALBAV for configurations with DATM and MPASO. It was incorrectly set in CIME5 development. It will change answers for C and G cases. However, 10-year tests of -compset GMPAS-IAF -mach anvil -res T62_oEC60to30v3 with this variable true and false show the effect is not climate-changing. [non-BFB] 01 August 2017, 18:36:31 UTC
bbe604b Junqi Yin 01 August 2017, 16:29:46 UTC Merge branch 'bishtgautam/lnd/fix-namelist-build' (PR #1669) Updates ALM build script for BGC relatved variables Values for suplnitro and suplphos from the user defined namelist are used during the generation of land namelist file. Adds few more consistency checks. Removes error checking for CTC and CNT soil bgc models. The error checking needs to be implemented in setup_cmdl_bgc(), which will be updated in the future. Fixes #1565 [BFB] 01 August 2017, 16:29:46 UTC
ee4d982 Jayesh Krishna 01 August 2017, 15:58:53 UTC Merge branch 'lcrc/machines/bebop' into next (PR #1662) Adding support for ANL Bebop cluster. [BFB] Conflicts: cime/config/acme/machines/config_batch.xml 01 August 2017, 15:58:53 UTC
70626e9 Azamat Mametjanov 01 August 2017, 15:39:25 UTC Merge branch 'alcf/machines/theta' (PR #1689) Update ALCF Theta machine config [BFB] 01 August 2017, 15:39:25 UTC
ef3a712 Azamat Mametjanov 01 August 2017, 15:35:31 UTC Merge branch 'alcf/machines/theta' into next (PR #1689) Update ALCF Theta machine config [BFB] 01 August 2017, 15:35:31 UTC
a6a42ba Azamat Mametjanov 01 August 2017, 15:16:46 UTC Merge branch 'azamat/homme/fix-io-init' (PR #1684) Initialize HOMME netcdf file open var: netcdf_io_mod::nf_open_file routine does not initialize ncFileID, which is used by other routines to drive MPI calls. All IO tasks should have a non-negative value for ncFileID (e.g. masterproc has ncFileID=0) to avoid hangs. This PR sets it to 0 for all tasks that call nf_open_file. Also, this PR updates Mira/Cetus machine config: * link to correct HDF5 and Zlib libraries * use pre-installed cprnc, instead of building a new one * suppress nested-omp compiler warnings. [BFB] Fixes #1683 01 August 2017, 15:16:46 UTC
8aad11e Azamat Mametjanov 31 July 2017, 21:59:41 UTC Move Bebop baselines to a subdirectory 31 July 2017, 21:59:41 UTC
fe498fb Azamat Mametjanov 31 July 2017, 21:28:47 UTC Merge branch 'azamat/homme/fix-io-init' into next (PR #1684) Initialize HOMME netcdf file open var: netcdf_io_mod::nf_open_file routine does not initialize ncFileID, which is used by other routines to drive MPI calls. All IO tasks should have a non-negative value for ncFileID (e.g. masterproc has ncFileID=0) to avoid hangs. This PR sets it to 0 for all tasks that call nf_open_file. Also, this PR updates Mira/Cetus machine config: * link to correct HDF5 and Zlib libraries * use pre-installed cprnc, instead of building a new one * suppress nested-omp compiler warnings. [BFB] Fixes #1683 31 July 2017, 21:28:47 UTC
c249fdf Gautam Bisht 31 July 2017, 21:17:40 UTC Adds I-Compsets for 20th-century with RTM Adds following four compsets: - I20TRCNECACTCBC : Carbon-Nitrogen with Century soil BGC + RTM - I20TRCNECACNTBC : Carbon-Nitrogen with CTC soil BGC + RTM - I20TRCNPECACTCBC : Carbon-Nitrogen-Phosphorus with Century soil BGC + RTM - I20TRCNPECACNTBC : Carbon-Nitrogen-Phosphorus with CTC soil BGC + RTM 31 July 2017, 21:17:40 UTC
349a0de Azamat Mametjanov 31 July 2017, 20:45:46 UTC Update batch settings 31 July 2017, 20:45:46 UTC
a4f3eff Azamat Mametjanov 31 July 2017, 19:52:00 UTC Merge branch 'alcf/mira/pes-f-compset' (PR #1688) Update PEs for ne120 F-compset cases on Mira [BFB] 31 July 2017, 19:52:00 UTC
f55cf73 Azamat Mametjanov 31 July 2017, 19:49:40 UTC Merge branch 'alcf/mira/pes-f-compset' into next (PR #1688) Update PEs for ne120 F-compset cases on Mira [BFB] 31 July 2017, 19:49:40 UTC
f4890bc Azamat Mametjanov 31 July 2017, 19:40:35 UTC Update PEs for ne120 F-compset cases on Mira 31 July 2017, 19:40:35 UTC
aa740da Azamat Mametjanov 31 July 2017, 18:11:34 UTC Update default project on Theta 31 July 2017, 18:11:34 UTC
c9b1e61 Matthew Hoffman 31 July 2017, 17:33:11 UTC Update MPASLI to fix bit reproducibility The previous update of MPASLI in this branch brought in changes that were not BFB for the configuration used in ACME. In fixing that bug, I also discovered that the configuration in ACME was not bit-reproducible. So this update also fixes that bug. However, that fix makes this update answer-changing. Differences are small, and the new results are identical across different processor counts. I also update the namelist option to force the Albany mesh to be rebuilt on every time step. This is needed to allow Albany runs to be BFB-restartable until a fix in Albany is brought into ACME. [Non-BFB] 31 July 2017, 17:39:39 UTC
436d07c Azamat Mametjanov 29 July 2017, 22:25:58 UTC Resolve paths from $PROJECT 29 July 2017, 22:25:58 UTC
74f9135 Azamat Mametjanov 29 July 2017, 18:30:15 UTC Update Theta machine config 29 July 2017, 18:30:15 UTC
b644bcc Junqi Yin 29 July 2017, 18:00:28 UTC Merge branch 'bishtgautam/lnd/fix-namelist-build' into next (PR #1669) Updates ALM build script for BGC relatved variables Values for suplnitro and suplphos from the user defined namelist are used during the generation of land namelist file. Adds few more consistency checks. Removes error checking for CTC and CNT soil bgc models. The error checking needs to be implemented in setup_cmdl_bgc(), which will be updated in the future. Fixes #1565 [BFB] 29 July 2017, 18:04:27 UTC
9d6361f Gautam Bisht 29 July 2017, 06:38:07 UTC Fixes setup of land namelist for soil BGC Removes error checking for CTC and CNT soil bgc models. The error checking needs to be implemented in setup_cmdl_bgc(), which will be updated in the future. 29 July 2017, 06:38:07 UTC
5f7fdda Gautam Bisht 29 July 2017, 01:55:51 UTC Merge branch 'jinyuntang/lnd/betr_v2int' into next (PR #1618) This replaces betr-v1 with the more modularized betr-v2. The code run successfully with gnu compiler on edison, however, it requires intel 'compiler 16.x to run and 15.x can only run in the debug model. This replacement does not affect other parts of ALM other than betr. [BFB] 29 July 2017, 01:55:51 UTC
1855bcb Azamat Mametjanov 28 July 2017, 21:16:50 UTC Perl module is not installed, use a local one 28 July 2017, 21:16:50 UTC
d13eadf Azamat Mametjanov 28 July 2017, 20:34:31 UTC Fix module cmd when called from python 28 July 2017, 20:34:31 UTC
44201fe Jinyun Tang 28 July 2017, 20:10:23 UTC bug fix for ibm compiler Add a logical switch to workaround the failure of creating the betr_parameter namelist on the fly. No change to ALM is made. 28 July 2017, 20:10:23 UTC
7a40529 Junqi Yin 28 July 2017, 18:24:05 UTC Merge branch 'bishtgautam/lnd/fix-namelist-build' into next (PR #1669) Updates ALM build script for BGC relatved variables Values for suplnitro and suplphos from the user defined namelist are used during the generation of land namelist file. Adds few more consistency checks. Fixes #1565 [BFB] 28 July 2017, 18:24:05 UTC
d963c03 noel 28 July 2017, 16:22:24 UTC Merge branch 'ndk/machinefiles/cori-knl-intelmpi-intel18b' into next (PR #1685) Add two new compiler options for cori-knl to make it easier for testing For cori-knl only right now: --compiler=intel18 to use the beta version of Intel v18 --compiler=intel-impi to use the Intel MPI (with the default Intel compiler) [BFB] 28 July 2017, 16:22:24 UTC
40b6535 noel 28 July 2017, 15:55:21 UTC Correct and improve the intel compiler flags for trying intel18 (version 18 of Intel compiler) and the option to use Intel MPI 28 July 2017, 15:55:21 UTC
879fd24 Gautam Bisht 26 July 2017, 19:53:59 UTC Adds a land test for 20th century transient A test with active land BGC for the 20th century transient is added. This test will ensure the dynamic PFT capability in ALM will be protected. 28 July 2017, 15:10:34 UTC
2adbce1 Junqi Yin 28 July 2017, 14:05:54 UTC Change clm-interface's zi dimension to be consistent with ALM [BFB] 28 July 2017, 14:05:54 UTC
ff93994 Junqi Yin 28 July 2017, 14:00:54 UTC Add a new test for ALM-PFLOTRAN's bgc interface [BFB] 28 July 2017, 14:00:54 UTC
bd5ddeb noel 28 July 2017, 00:47:05 UTC Merge branch ACME-Climate/PeterCaldwell/cime/add-edison-Fcompset-PElayout (PR #1637) Default PE layout for F-compsets on Edison 28 July 2017, 00:47:05 UTC
ef3387f noel 28 July 2017, 00:22:48 UTC Merge branch 'PeterCaldwell/cime/add-edison-Fcompset-PElayout' into next (PR #1637) Currently, unless you explicitly specify a custom PE layout on Edison, F-compset runs will default to 192 tasks + 4 threads (=32 total nodes), yielding <2 SYPD. This PR specifies a reasonable default for ne30 F compsets. Note that what is proposed here isn't an optimal layout. Hyperthreading can increase SYPD by 1 or more, but it isn't turned on by default any more. This layout was chosen simply because it works and isn't horrible. just allows newbies or people who forget to specify the PE layout to get something that isn't totally useless. Efforts to get a faster PE layout would be useful. Note that this is a second attempt to do this. #1241 also proposed a similar F PE layout but failed to pass tests for unrelated issues documented here: #1263 (measured 10.9 sypd after edison upgrade) 28 July 2017, 00:22:48 UTC
26276c6 Azamat Mametjanov 28 July 2017, 00:14:04 UTC Suppress nested-omp IBM compiler warnings 28 July 2017, 00:14:04 UTC
d90038f Azamat Mametjanov 28 July 2017, 00:11:46 UTC Init HOMME internal netcdf control-var Also update Mira/Cetus machine config: * link to correct HDF5 and Zlib * use pre-installed cprnc * suppress nested-omp compiler warnings 28 July 2017, 00:11:46 UTC
f7bbe57 noel 27 July 2017, 23:48:20 UTC For cori-knl only: Add option to compile with intel18 beta version. --compiler=intel18 Add option to build with Intel MPI. --compiler=intel-impi To use Intel MPI, can't use Cray compiler wrappers and so need to add some flags manually 27 July 2017, 23:48:20 UTC
a4aa68e Michael Deakin 27 July 2017, 23:27:22 UTC Merge branch 'mfdeakin-sandia/run_acme/submission_cleanup' 27 July 2017, 23:27:22 UTC
c572458 Michael Deakin 27 July 2017, 23:25:32 UTC Merge branch 'mfdeakin-sandia/run_acme/submission_cleanup' into next 27 July 2017, 23:25:32 UTC
98ba7c8 Gautam Bisht 27 July 2017, 22:59:19 UTC Merge branch 'jinyuntang/lnd/betr_v2int' into next (PR #1618) This replaces betr-v1 with the more modularized betr-v2. The code run successfully with gnu compiler on edison, however, it requires intel 'compiler 16.x to run and 15.x can only run in the debug model. This replacement does not affect other parts of ALM other than betr. [BFB] 27 July 2017, 22:59:19 UTC
5a6abd0 Michael Deakin 27 July 2017, 20:54:31 UTC Minor cleanups for the run_acme submission code 27 July 2017, 22:57:32 UTC
bcaa479 Jinyun Tang 27 July 2017, 22:43:20 UTC update sbetr with bug fixes Now sbetr standalone tests passed. No change to ALM is involved. 27 July 2017, 22:43:20 UTC
1efdd70 Wade Burgess 27 July 2017, 22:12:37 UTC Merge branch 'pwburge/cime/chama-WCID-change' into master (PR #1682) Updated WCID for chama Updated the config_machines.xml file to edit chama's tag, which is the WCID for ACME on chama. [BFB] 27 July 2017, 22:16:10 UTC
f9dd5c5 Wade Burgess 27 July 2017, 22:12:37 UTC Merge branch 'pwburge/cime/chama-WCID-change' into next (PR #1682) Updated WCID for chama Updated the config_machines.xml file to edit chama's tag, which is the WCID for ACME on chama. [BFB] 27 July 2017, 22:12:37 UTC
3964231 noel 27 July 2017, 22:12:05 UTC Merge branch ACME-Climate/ndk/machinefiles/edison-2017-update-hotfix (PR #1679) Update to support edison after OS/software upgrades 27 July 2017, 22:12:05 UTC
546e693 James Foucar 27 July 2017, 21:09:30 UTC Merge branch 'azamat/mem-usage/swap-highwater' into next (PR #1639) Swap mem highwater and usage logs for memleak tests This is needed for correct parsing of mem highwater values from cpl.log. Previously, memory resident set size was getting parsed while checking for memleaks. Fixes #1636 [BFB] * origin/azamat/mem-usage/swap-highwater: Swap mem highwater and usage logs for memleak tests 27 July 2017, 21:09:30 UTC
dd4151b Wade Burgess 27 July 2017, 21:00:42 UTC Updated WCID for chama Updated the config_machines.xml file to edit chama's <PROJECT> tag, which is the WCID for ACME on chama. [BFB] 27 July 2017, 21:00:42 UTC
b4dadaa noel 27 July 2017, 20:48:55 UTC Merge branch 'ndk/machinefiles/edison-2017-update-hotfix' into next (PR #1679) Edison back after a OS/software upgrade that makes it look more like Cori. All codes have to recompile. Almost all ACME modules needed to be updated, including going to Intel v17 as default (v15 no longer avail). However, the same netcdf/hdf versions used previously are still there, so using those for now. The -c flag to srun now also needs to be present. When using 24 MPI's per node, it should be -c 2 --cpu_bind=cores. Added some logic to set this automatically for us. Removed the intel17 compiler option (as well as cray compiler that we were not testing). Cleaned up entries in config_compilers.xml for all 3 NERSC machines and made then look more similar. This does affect minor build details with cori-haswell and cori-knl (ie removing -lm from linking and a flag from debug build), but I'm also testing there as well. Tested with acme_developer using Intel/GNU and the results were the same as before the upgrade (with intel17). Verified good performance with ne30 F case (19.2 sypd on 225 nodes) 27 July 2017, 20:48:55 UTC
0574899 noel 27 July 2017, 20:24:42 UTC Remove comment. 27 July 2017, 20:24:42 UTC
1c26e7a noel 27 July 2017, 20:18:52 UTC Minor cleanup. After removing that flag, can now simplify and inherit those flags from the Intel stanza above. 27 July 2017, 20:18:52 UTC
91b0cac noel 27 July 2017, 20:13:13 UTC Correcting a mistake in my last commit -- I meant to remove the -qno-opt-dynamic-align flag from the DEBUG build. We want it for optimized builds, but shouldn't need it for DEBUG. 27 July 2017, 20:13:13 UTC
446fbf1 noel 27 July 2017, 19:58:23 UTC Changes after review. Remove the -lm flag as Intel compiler recommends. Further simplification in config_compilers as we use what is defined with CNL. Prefer shell commands over using python. 27 July 2017, 19:58:23 UTC
back to top