Search found 11 matches
- Thu Sep 12, 2019 6:36 pm
- Forum: User discussion
- Topic: Passing magma_queue_t to magma_?trsm
- Replies: 0
- Views: 1321
Passing magma_queue_t to magma_?trsm
Hi, I've been trying to use the magma_dtrsm routine, but it doesn't work as documented here: http://icl.cs.utk.edu/projectsfiles/magma/doxygen/group__magma__trsm.html#ga36b2b66d52f384db19cf256cc5e8a4c2. It expects a magma_queue_t, but when I pass one, I get errors telling me that I'm passing too man...
- Fri Sep 06, 2019 3:59 pm
- Forum: User discussion
- Topic: (d/s)getrf_batched doesn't send back errors
- Replies: 4
- Views: 2444
Re: (d/s)getrf_batched doesn't send back errors
Hi Ahmad, Just pinging back about this issue. The original issue has been resolved, but I find a minor inconsistency between the returned values of the getrf and getrf_batched. In getrf, if the input is singular, the diagonal of the returned matrix has 0s, whereas in getrf_batched, if the inputs is ...
- Sat Jul 13, 2019 11:59 pm
- Forum: User discussion
- Topic: [Bug?] spotrf_batched fails for batch size > 524280 and dpotrf_batched fails for batch size > 262140
- Replies: 1
- Views: 1725
[Bug?] spotrf_batched fails for batch size > 524280 and dpotrf_batched fails for batch size > 262140
Hi, I found out that the spotrf_batched routine fails for batch size > 524280 and dpotrf_batched routine fails for batch size > 262140. I confirmed this issue using the test suite and the following commands: ./testing_spotrf_batched --batch 524281 -n 2 --check --matrix rand_dominant and ./testing_dp...
- Sun May 12, 2019 11:42 pm
- Forum: User discussion
- Topic: [BUG, Fixed] (d/s)potrf_batched doesn't send back errors
- Replies: 4
- Views: 2470
Re: [BUG] (d/s)potrf_batched doesn't send back errors
Hi Ahmad,
Thank you for fixing this. I can confirm that it reports the correct error code now.
Best,
Vishwak
Thank you for fixing this. I can confirm that it reports the correct error code now.
Best,
Vishwak
- Wed May 08, 2019 10:56 am
- Forum: User discussion
- Topic: [BUG, Fixed] (d/s)potrf_batched doesn't send back errors
- Replies: 4
- Views: 2470
Re: [BUG] (d/s)potrf_batched doesn't send back errors
Hi Ahmad, Thank you for the message. The specific case where I see that the error code is not sent back correctly is when you have a constant matrix. I verified that this is not working correctly by modifying the tests for magma_dpotrf_batched. When I initialize the matrix with 1 at every entry, the...
- Wed Feb 06, 2019 6:55 am
- Forum: User discussion
- Topic: [BUG, Fixed] (d/s)potrf_batched doesn't send back errors
- Replies: 4
- Views: 2470
[BUG, Fixed] (d/s)potrf_batched doesn't send back errors
Hi, I'm using MAGMA to perform Cholesky decomposition on batches of matrices. While passing the batched routine matrices that are not positive definite, it doesn't seem to be passing back the errors as expected. However, the single batch version of this routine (d/spotrf_gpu) does it correctly for i...
- Sat Dec 29, 2018 5:16 am
- Forum: User discussion
- Topic: (d/s)potrf_batched has some kind of memory leak
- Replies: 2
- Views: 1177
Re: (d/s)potrf_batched has some kind of memory leak
Thank you for looking into this. I can verify that the bug has been fixed in the latest master.
- Fri Dec 28, 2018 9:03 am
- Forum: User discussion
- Topic: (d/s)potrf_batched has some kind of memory leak
- Replies: 2
- Views: 1177
(d/s)potrf_batched has some kind of memory leak
Hello, I seem to have discovered a memory leak in the potrf_batched routines when running the tests for the same routine. The GPU memory used increases with the number of iterations, which is uncharacteristic of other tests like getri, getrf and potrs. Could someone take a look at this? Thank you in...
- Wed Dec 12, 2018 11:18 am
- Forum: User discussion
- Topic: (d/s)getrf_batched doesn't send back errors
- Replies: 4
- Views: 2444
Re: (d/s)getrf_batched doesn't send back errors
Hi Ahmad, Thank you for the prompt response. Unfortunately, this doesn't seem to resolve the issue, and now the non-pivoting variant is also not working as expected with respect to sending back errors. Furthermore, I think I might have found that the pivot information is also not sent back in the pi...
- Mon Dec 10, 2018 2:43 am
- Forum: User discussion
- Topic: (d/s)getri_outofplace_batched can't handle many matrices
- Replies: 2
- Views: 1037
(d/s)getri_outofplace_batched can't handle many matrices
Hi, Sorry about the ambiguous title. I filed an issue a month ago about (d/s)getri_outofplace_batched not being able to handle more than 65535 matrices in a batch. This is the link to the issue: https://bitbucket.org/icl/magma/issues/10/getri_outofplace_batched-fails-when I will present the details ...