search

geodynamics / vq

Virtual Quake is a boundary element code designed to investigate long term fault system behavior and interactions between faults through stress transfer.
Other
12 stars 24 forks source link

Fix MPP Segmentation Fault and Linux installation #69

Closed markyoder closed 9 years ago

markyoder commented 9 years ago

1) minor change to quakelib:QuakeLibUtil.h "include" to facilitate Linux installations (use relative include path, not absolute (mac) path). 2) several memory management, and variable initialization, and some minor bug fixes that seem to be causing instability in MPI mode.

eheien commented 9 years ago

Could you remove some of the debugging code that's commented out, and run 'astyle --options=astyle.rc --recursive ".cpp" --recursive ".h' to make the formatting consistent? Thanks!

markyoder commented 9 years ago

ran astyle and pushed to my repo. i'll also have a look for debugging comments in master.

kwschultz commented 9 years ago

Is this good to go?

markyoder commented 9 years ago

let me go through it one more time. i've started some development since then, but i think it's all been on a new branch.

m

On Mon, Mar 23, 2015 at 1:51 PM, Kasey Schultz notifications@github.com wrote:

Is this good to go?

— Reply to this email directly or view it on GitHub https://github.com/geodynamics/vq/pull/69#issuecomment-85190637.

Mark Yoder, PhD 805 451 8750

"If you bore me, you lose your soul..." ~belly

markyoder commented 9 years ago

kasey -

this looks good to go from my perspective. as per eric's request, there is probably debugging commentary yet to be removed, but since we're still chasing down a couple of other bugs and since these changes are fresh, i'd assume leave them in for now. as we go forward and these fixes are confirmed (and so the comments are no longer relevant), we'll remove them.

tentatively, i think we're a few days away from another pull request to incorporate these final_stress nan values, but there is evaluation yet to be done there.

if you get a chance, you might run your own tests before approving the merge. i think the criteria for this scope of work is that 1) it runs and is stable in MPI mode, 2) debugging lines don't print to screen, 3) we get output files (though the nan problem remains).

thanks.

m

On Mon, Mar 23, 2015 at 2:02 PM, Mark Yoder mark.yoder@gmail.com wrote:

let me go through it one more time. i've started some development since then, but i think it's all been on a new branch.

m

On Mon, Mar 23, 2015 at 1:51 PM, Kasey Schultz notifications@github.com wrote:

Is this good to go?

— Reply to this email directly or view it on GitHub https://github.com/geodynamics/vq/pull/69#issuecomment-85190637.

Mark Yoder, PhD 805 451 8750

"If you bore me, you lose your soul..." ~belly

Mark Yoder, PhD 805 451 8750

"If you bore me, you lose your soul..." ~belly

markyoder commented 9 years ago

still getting a "can't automatically merge" message, so i assume there are some conflicts which i'll get sorted out asap.

markyoder commented 9 years ago

a new, much improved pull request will be generated with many, many issues resolved.