Bringing back the new gradient calculation to the master branch #35
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
There are still some naughty comments in there and a problem of how to restore zb in the halo!
src/ConserveElevation.cu
Outdated
| @@ -852,6 +868,8 @@ template <class T> void conserveElevationLeft(Param XParam,int ib, int ibLB, int | |||
| // Restriction | |||
| ihalo = -1; | |||
| ip = XParam.blkwidth - 2; | |||
|
|
|||
| int ii = memloc(XParam, -1, 6, 46); | |||
src/ConserveElevation.cu
Outdated
| @@ -863,6 +881,8 @@ template <class T> void conserveElevationLeft(Param XParam,int ib, int ibLB, int | |||
| } | |||
|
|
|||
| } | |||
| if (ib == 46) | |||
| printf("Restriction1: hii=%f, \n", XEv.h[ii]); | |||
There was a problem hiding this comment.
Debugging stuff remove
src/ConserveElevation.cu
Outdated
| @@ -874,7 +894,8 @@ template <class T> void conserveElevationLeft(Param XParam,int ib, int ibLB, int | |||
| } | |||
|
|
|||
| } | |||
|
|
|||
| if (ib == 46) | |||
| printf("Restriction2: hii=%f, \n", XEv.h[ii]); | |||
src/ConserveElevation.cu
Outdated
|
|
||
| ProlongationElevation(XParam.halowidth, XParam.blkmemwidth, T(XParam.eps), ib, ibn, ihalo, jhalo, il, jl, ip, jp, XEv.h, XEv.zs, zb); | ||
| } | ||
|
|
||
| } | ||
| if (ib == 46) | ||
| printf("Prolongation: hii=%f, \n", XEv.h[ii]); |
src/Halo.cu
Outdated
| @@ -115,6 +115,8 @@ template void fillHaloTopRightGPU<float>(Param XParam, BlockP<float> XBlock, cud | |||
| template <class T> void fillHalo(Param XParam, BlockP<T> XBlock, EvolvingP<T> Xev, T*zb) | |||
| { | |||
|
|
|||
| int ii = memloc(XParam, -1, 6, 46); | |||
| printf("prehalo: h[ii]=%f, zb[ii]=%f\n", Xev.h[ii], zb[ii]); | |||
|
|
||
| printf("h[i]=%f; h[ileft]=%f dhdx[i]=%f, dhdx[ileft]=%f, zs[i]=%f, zs[ileft]=%f, dzsdx[i]=%f, dzsdx[ileft]=%f\n", XModel.evolv.h[i], XModel.evolv.h[ileft], XModel.grad.dhdx[i], XModel.grad.dhdx[ileft], XModel.evolv.zs[i], XModel.evolv.zs[ileft], XModel.grad.dzsdx[i], XModel.grad.dzsdx[ileft]); |
There was a problem hiding this comment.
No this serves a debbug purpose in the test!
src/ConserveElevation.cu
Outdated
| @@ -7,10 +7,18 @@ template <class T> void conserveElevation(Param XParam, BlockP<T> XBlock, Evolvi | |||
| for (int ibl = 0; ibl < XParam.nblk; ibl++) | |||
| { | |||
| ib = XBlock.active[ibl]; | |||
|
|
|||
| int ii = memloc(XParam, -1, 5, 46); | |||
src/ConserveElevation.cu
Outdated
| @@ -852,6 +887,8 @@ template <class T> void conserveElevationLeft(Param XParam,int ib, int ibLB, int | |||
| // Restriction | |||
| ihalo = -1; | |||
| ip = XParam.blkwidth - 2; | |||
|
|
|||
| int ii = memloc(XParam, -1, 5, 46); | |||
| top = memloc(XParam.halowidth, XParam.blkmemwidth, n, XParam.blkwidth, ib); | ||
| bot = memloc(XParam.halowidth, XParam.blkmemwidth, n, -1, ib); | ||
|
|
||
| Xev.zs[left] = zb[left] + Xev.h[left]; |
There was a problem hiding this comment.
Why is this doing anything ifferent than the GPU? Is this function called by anyone?
There was a problem hiding this comment.
There is reason think that the recalculate zs should only apply in teh halo. I will remove to be consistent with GPOU side but this needs to be double checked to reduce computational demand.
|
|
||
| RecalculateZs(XParam, XBlock, Xev, zb); |
There was a problem hiding this comment.
This will lead to different results as the GPU version!
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
WTF is going on with the Recalculatezs functions where the CPU and GPU version seem to be doing very different things. Is that a debugging left-over mess? Needs double checking before merging.