| 1 |
Hi, |
| 2 |
|
| 3 |
*Summary:* |
| 4 |
1) Continue work on exported functions and respective environment variables. |
| 5 |
2) Start to objdump (among other things) to try and link against only |
| 6 |
MPI-dependent libs. |
| 7 |
3) Continue ironing out bugs and errors from build.log. |
| 8 |
|
| 9 |
The build.log is looking much better now. I had to travel this week so some |
| 10 |
other objectives were neglected- I will be making up for this next week |
| 11 |
when I will get to sit still for a bit :) |
| 12 |
|
| 13 |
Next week's plans: |
| 14 |
1) Make helper functions more comprehensive with regards to setting |
| 15 |
environment variables. Modularize functions to help in more cases than one. |
| 16 |
2) Finish work with setting up a proper build environement for the src_* |
| 17 |
phases. |
| 18 |
3) Test against new version of hpl from sci/ now in the overlay, along with |
| 19 |
new MPI_TARGETS. |
| 20 |
|
| 21 |
*Day-by-day breakdown:* |
| 22 |
|
| 23 |
|
| 24 |
*Monday plans:* |
| 25 |
|
| 26 |
Expand upon mpi_foreach. |
| 27 |
Work in progress. mpi_src_install_all will need populating. |
| 28 |
|
| 29 |
Correct repoman errors until repoman is partially happy. |
| 30 |
Completed. There is more work to be done with regards to repoman, but now |
| 31 |
it will accept the new changes. |
| 32 |
|
| 33 |
|
| 34 |
*Tuesday Plans:* |
| 35 |
|
| 36 |
Today my flight got delayed until very late at night, and I did not have |
| 37 |
access to my testing environment at the airport. Work on this will continue |
| 38 |
tomorrow- I will also comment on another email thread shortly. |
| 39 |
|
| 40 |
|
| 41 |
*Wednesday Plans:* |
| 42 |
|
| 43 |
*Correct command not found errors from repoman.* |
| 44 |
|
| 45 |
Work in progress. Unfortunately, my testing environment I was using is no |
| 46 |
longer functioning. More specifically: previously I was dependent on having |
| 47 |
the charger plugged in due to battery malfunctions, but now the charger is |
| 48 |
no longer working, so I cannot boot up the machine. I've been working on |
| 49 |
getting a new VM environment up to properly test on. |
| 50 |
|
| 51 |
|
| 52 |
*Update function for RDEPEND contents.* |
| 53 |
Incomplete. I will be working on this tomorrow. |
| 54 |
|
| 55 |
|
| 56 |
|
| 57 |
*Thursday Plans:* |
| 58 |
|
| 59 |
Get the VM envrionment up-to-speed and ready to test on. |
| 60 |
|
| 61 |
Completed. This should (hopefully) be the solution until the end of summer |
| 62 |
:) |
| 63 |
|
| 64 |
|
| 65 |
Continue work on repoman errors. Even though the repoman check will "pass" |
| 66 |
these seem to be critical. |
| 67 |
|
| 68 |
Thanks to Justin's insight this is finally complete! I need to follow up |
| 69 |
about this, though. |
| 70 |
|
| 71 |
|
| 72 |
Expand upon mpi_dependencies function. |
| 73 |
|
| 74 |
Incomplete. I was having some technical difficulties last night with our |
| 75 |
hotels wifi, which is also why this email is so terribly late, so I will be |
| 76 |
working on this Friday. |
| 77 |
|
| 78 |
|
| 79 |
*Friday's Plans:* |
| 80 |
|
| 81 |
Expand upon mpi_dependencies function. |
| 82 |
|
| 83 |
Work in progress. While it still needs to be more comprehensive and include |
| 84 |
more from mpi_pkg_deplist, the space checking should be a small improvement. |
| 85 |
|
| 86 |
|
| 87 |
Test against new MPI_TARGETS, or other hpl version in sci/ and add to the |
| 88 |
overlay respectively. |
| 89 |
|
| 90 |
I've added the 2nd version of hpl from sci, but I need to test against it |
| 91 |
more and modify mpi-select accordingly. |
| 92 |
|
| 93 |
Thanks, |
| 94 |
Michael |
Archives