search

danigeos / tisc

TISC is a code to model isostasy, erosion/sediment transport/sedimentation, and tectonic kinematics, mostly for the study of sedimentary basins and source to sink systems.
https://sites.google.com/site/daniggcc/software/tisc
16 stars 2 forks source link

Seg Fault with high number of UNIT files #25

Open mberry28 opened 2 years ago

mberry28 commented 2 years ago

Hi Daniel,

We have several unit files we are using to simulate progressive subsidence using the 'hidden' command. We hit a point that when we added an additional UNIT file we get a segmentation fault after a few tectonic time steps. I have removed the 'hidden' blocks and the code runs normally. I am wondering if you have any suggestions. Is it possible to allocate more ram allowance to TISC? after some checks it seems like it doesn't even gets close to capping out the ram. Any other solutions?

Thanks

danigeos commented 2 years ago

How many units / blocks are you defining? Note that 'hidden' blocks are not even registered, no need for that since they do not have topographic expression except for the isostatic response they imply. So they should not make memory problems.

On Fri, Apr 29, 2022 at 6:52 AM Michael Berry @.***> wrote:

Hi Daniel,

We have several unit files we are using to simulate progressive subsidence using the 'hidden' command. We hit a point that when we added an additional UNIT file we get a segmentation fault after a few tectonic time steps. I have removed the 'hidden' blocks and the code runs normally. I am wondering if you have any suggestions, here are what I think are possible solutions: Is it possible to allocate more ram allowance to TISC? after some checks it seems like it doesn't even gets close to capping out the ram. Any other solutions?

Thanks

— Reply to this email directly, view it on GitHub https://github.com/danigeos/tisc/issues/25, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACYFBCSPD4OS6SJVTH3NXRTVHNTKPANCNFSM5UUUAPPQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

mberry28 commented 2 years ago

There are 14 UNIT files, 6 of which are hidden blocks. If I drop on of the hidden loads (to 5) the run progress fine with no Segmentation Fault. If I add in the additional block the Seg fault returns.

I also adjusted the timing of the block. if time_start is at 8, then the seg fault happens at 8 Myr, if I change time_start to 9, the seg fault happens at 9. So it seems to be related.

I also tried to have, instead of all the hidden UNIT files being active at the same time, I have them being active for only one tectonic time step, this also causes the seg fault. Hope I am making sense here.

danigeos commented 2 years ago

Got it! The zip file with the project was needed.

Just download the new version, it's complicated to explain. Error related to freeing an array that is not always defined...

Note: your PRFL is not within the x limits. the time for unit 9 is wrong, remember, UNIT 'time' must increase with the UNIT number

Anyway, your project is working right now. Let me know otherwise let me know

On Tue, May 3, 2022 at 7:14 AM Michael Berry @.***> wrote:

There are 14 UNIT files, 6 of which are hidden blocks. If I drop on of the hidden loads (to 5) the run progress fine with no Segmentation Fault. If I add in the additional block the Seg fault returns.

I also adjusted the timing of the block. if time_start is at 8, then the seg fault happens at 8 Myr, if I change time_start to 9, the seg fault happens at 9. So it seems to be related.

I also tried to have, instead of all the hidden UNIT files being active at the same time, I have them being active for only one tectonic time step, this also causes the seg fault. Hope I am making sense here.

— Reply to this email directly, view it on GitHub https://github.com/danigeos/tisc/issues/25#issuecomment-1115746321, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACYFBCS3JJRZ36UVKWHXJTLVICY23ANCNFSM5UUUAPPQ . You are receiving this because you commented.Message ID: @.***>

mberry28 commented 2 years ago

Ah yes.

I am sure that it was the UNIT files not being in the right time! I will test it out on my end to verify.

I will update to the newest TISC too... thought I was using the newest.

Thank you.

On Wed, May 4, 2022 at 8:28 AM Daniel Garcia-Castellanos < @.***> wrote:

Got it! The zip file with the project was needed.

Just download the new version, it's complicated to explain. Error related to freeing an array that is not always defined...

Note: your PRFL is not within the x limits. the time for unit 9 is wrong, remember, UNIT 'time' must increase with the UNIT number

Anyway, your project is working right now. Let me know otherwise let me know

On Tue, May 3, 2022 at 7:14 AM Michael Berry @.***> wrote:

There are 14 UNIT files, 6 of which are hidden blocks. If I drop on of the hidden loads (to 5) the run progress fine with no Segmentation Fault. If I add in the additional block the Seg fault returns.

I also adjusted the timing of the block. if time_start is at 8, then the seg fault happens at 8 Myr, if I change time_start to 9, the seg fault happens at 9. So it seems to be related.

I also tried to have, instead of all the hidden UNIT files being active at the same time, I have them being active for only one tectonic time step, this also causes the seg fault. Hope I am making sense here.

— Reply to this email directly, view it on GitHub https://github.com/danigeos/tisc/issues/25#issuecomment-1115746321, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ACYFBCS3JJRZ36UVKWHXJTLVICY23ANCNFSM5UUUAPPQ

. You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/danigeos/tisc/issues/25#issuecomment-1117558480, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADFJZKBRR5WOOWKMZNM6YZDVIKQTNANCNFSM5UUUAPPQ . You are receiving this because you authored the thread.Message ID: @.***>