Venado optimizations #755

mewall · 2025-01-14T16:54:26Z

Modify bml_transpose() fortran API to match the C API
o Add bml_transpose_new()
o Change the tests to use the new API
Add methods to get the pointer for MAGMA arrays, for use in Fortran OpenACC and OpenMP offload
o Write fortran wrapper for existing bml_get_data_ptr_dense()
o Add new bml_get_ld_dense()
Add bml_set_N_dense() to change the size of a bml array that's already been allocated
o This avoids unnecessary allocations and leads to substantial speedups
o Unsafe method that's exposed in fortran for dense matrices only

o Back out bml_get_ptr_dense() from bml_getters o Write fortran wrapper for existing bml_get_data_ptr_dense() method o Write new bml_get_ld_dense() to enable magma matrix pointer use

nicolasbock

LGTM

scripts/build_bml_cray.sh

src/C-interface/dense/bml_setters_dense.c

scripts/build_chicoma_hackathon.sh

mewall · 2025-01-28T00:41:44Z

Yes, I think such a function could be good. The dense case is essentially done, even for magma build, we’d just need to add the allocated size to the struct and add a check to make it work. Need to figure out the other matrix types, if they’ll be supported. Meanwhile, what do you think about merging the current function? Get Outlook for iOS<https://aka.ms/o0ukef>

________________________________ From: Jean-Luc Fattebert ***@***.***> Sent: Monday, January 27, 2025 4:21:48 PM To: lanl/bml ***@***.***> Cc: Wall, Michael E ***@***.***>; Author ***@***.***> Subject: [EXTERNAL] Re: [lanl/bml] Venado optimizations (PR #755) @jeanlucf22 commented on this pull request.

________________________________ In src/C-interface/dense/bml_setters_dense.c<https://urldefense.com/v3/__https://github.com/lanl/bml/pull/755*discussion_r1931312603__;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$>:

@@ -3,6 +3,22 @@

#include "bml_setters_dense.h" #include "bml_types_dense.h" +#ifdef BML_USE_MAGMA +#include "magma_v2.h" +#endif + +void bml_set_N_dense( I understand now. I was not suggesting to use domain and domain2. I was just saying their allocation may be the main culprit when it comes to allocation time for a dense matrix. Maybe an issue to deal with another time. Another suggestion: having a function resizeNoAlloc(int n) that would just change N if n<=N, otherwise would change N and reallocate memory? Having an extra struct member keeping track of allocated memory size would be good in that case. — Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/lanl/bml/pull/755*discussion_r1931312603__;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AA67VEIOEH32HB66PF4BJQ32M25QZAVCNFSM6AAAAABVFKYFIOVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDKNZWG42TCNBWGM__;!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8iWVuCBR$>. You are receiving this because you authored the thread.Message ID: ***@***.***>

jeanlucf22 · 2025-01-29T19:24:40Z

Yes, I think such a function could be good. The dense case is essentially done, even for magma build, we’d just need to add the allocated size to the struct and add a check to make it work. Need to figure out the other matrix types, if they’ll be supported. Meanwhile, what do you think about merging the current function? Get Outlook for iOShttps://aka.ms/o0ukef
________________________________ From: Jean-Luc Fattebert @.> Sent: Monday, January 27, 2025 4:21:48 PM To: lanl/bml @.> Cc: Wall, Michael E @.>; Author @.> Subject: [EXTERNAL] Re: [lanl/bml] Venado optimizations (PR #755) @jeanlucf22 commented on this pull request.
________________________________ In src/C-interface/dense/bml_setters_dense.chttps://urldefense.com/v3/__https://github.com/lanl/bml/pull/755*discussion_r1931312603__;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$:
@@ -3,6 +3,22 @@
#include "bml_setters_dense.h" #include "bml_types_dense.h" +#ifdef BML_USE_MAGMA +#include "magma_v2.h" +#endif + +void bml_set_N_dense( I understand now. I was not suggesting to use domain and domain2. I was just saying their allocation may be the main culprit when it comes to allocation time for a dense matrix. Maybe an issue to deal with another time. Another suggestion: having a function resizeNoAlloc(int n) that would just change N if n<=N, otherwise would change N and reallocate memory? Having an extra struct member keeping track of allocated memory size would be good in that case. — Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https://github.com/lanl/bml/pull/755*discussion_r1931312603__;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AA67VEIOEH32HB66PF4BJQ32M25QZAVCNFSM6AAAAABVFKYFIOVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDKNZWG42TCNBWGM__;!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8iWVuCBR$. You are receiving this because you authored the thread.Message ID: @.***>

I don't think it is a good idea to merge as is. That's what branches are for if you need it as is right away.

jmohdyusof · 2025-01-29T20:06:41Z

FWIW, for the ellpack format, you can probably make this work simply by making nnz(i) = 0 for i > n. Obviously M is only an upper bound already, and the loops would continue to run over all N rows, unless you decided to introduce another variable to truncate them.

mewall · 2025-01-29T21:31:29Z

Can you please provide a list of requirements for merging this into master?

…

________________________________ From: Jean-Luc Fattebert ***@***.***> Sent: Wednesday, January 29, 2025 12:25 PM To: lanl/bml ***@***.***> Cc: Wall, Michael E ***@***.***>; Author ***@***.***> Subject: [EXTERNAL] Re: [lanl/bml] Venado optimizations (PR #755) Yes, I think such a function could be good. The dense case is essentially done, even for magma build, we’d just need to add the allocated size to the struct and add a check to make it work. Need to figure out the other matrix types, if they’ll be supported. Meanwhile, what do you think about merging the current function? Get Outlook for iOShttps://aka.ms/o0ukef<https://urldefense.com/v3/__https://aka.ms/o0ukef__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47KSGeMqG$>

________________________________ From: Jean-Luc Fattebert @.> Sent: Monday, January 27, 2025 4:21:48 PM To: lanl/bml @.> Cc: Wall, Michael E @.>; Author @.> Subject: [EXTERNAL] Re: [lanl/bml] Venado optimizations (PR #755<https://urldefense.com/v3/__https://github.com/lanl/bml/pull/755__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47HtHaUCk$>) @jeanlucf22<https://urldefense.com/v3/__https://github.com/jeanlucf22__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47HKzc6_0$> commented on this pull request.

________________________________ In src/C-interface/dense/bml_setters_dense.chttps://urldefense.com/v3/__#755*discussion_r1931312603__;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$: @@ -3,6 +3,22 @@ #include "bml_setters_dense.h" #include "bml_types_dense.h" +#ifdef BML_USE_MAGMA +#include "magma_v2.h" +#endif + +void bml_set_N_dense( I understand now. I was not suggesting to use domain and domain2. I was just saying their allocation may be the main culprit when it comes to allocation time for a dense matrix. Maybe an issue to deal with another time. Another suggestion: having a function resizeNoAlloc(int n) that would just change N if n<=N, otherwise would change N and reallocate memory? Having an extra struct member keeping track of allocated memory size would be good in that case. — Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__#755*discussion_r1931312603__;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AA67VEIOEH32HB66PF4BJQ32M25QZAVCNFSM6AAAAABVFKYFIOVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDKNZWG42TCNBWGM__;!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8iWVuCBR$. You are receiving this because you authored the thread.Message ID: @.***> I don't think it is a good idea to merge as is. That's what branches are for if you need it as is right away. — Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/lanl/bml/pull/755*issuecomment-2622639216__;Iw!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47BQYtol7$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AA67VEK7MTI6M5EYAESKTYT2NETI7AVCNFSM6AAAAABVFKYFIOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRSGYZTSMRRGY__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47E1cI_bk$>. You are receiving this because you authored the thread.Message ID: ***@***.***>

mewall · 2025-01-29T21:47:30Z

I'll take a look, I think for ellpack it's possible that the bml_set_N() method could work as well, for resizing N in an existing matrix buffer. There could also be a bml_set_M() method for resizing M. I think the expectation should be that the data are lost upon resizing, and that there's no zeroing of elements, as the purpose is to make it fast by avoiding memory access.

…

________________________________ From: Jamal Mohd-Yusof ***@***.***> Sent: Wednesday, January 29, 2025 1:07 PM To: lanl/bml ***@***.***> Cc: Wall, Michael E ***@***.***>; Author ***@***.***> Subject: [EXTERNAL] Re: [lanl/bml] Venado optimizations (PR #755) FWIW, for the ellpack format, you can probably make this work simply by making nnz(i) = 0 for i > n. Obviously M is only an upper bound already, and the loops would continue to run over all N rows, unless you decided to introduce another variable to truncate them. — Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/lanl/bml/pull/755*issuecomment-2622730507__;Iw!!Bt8fGhp8LhKGRg!B99uRRq12z8IMCZGBgVUiQyDYd8GtmmNzTTbeYAvf0MbSkB75_z_8zxznk5Z9VDMvL1v_vGSlk4g39_XhAKe3P87$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AA67VEJ5TEFKXLLPCS5VK432NEYGRAVCNFSM6AAAAABVFKYFIOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRSG4ZTANJQG4__;!!Bt8fGhp8LhKGRg!B99uRRq12z8IMCZGBgVUiQyDYd8GtmmNzTTbeYAvf0MbSkB75_z_8zxznk5Z9VDMvL1v_vGSlk4g39_XhChlp8l8$>. You are receiving this because you authored the thread.Message ID: ***@***.***>

jeanlucf22 · 2025-01-30T03:02:01Z

Can you please provide a list of requirements for merging this into master?
…
________________________________ From: Jean-Luc Fattebert @.> Sent: Wednesday, January 29, 2025 12:25 PM To: lanl/bml @.> Cc: Wall, Michael E @.>; Author @.> Subject: [EXTERNAL] Re: [lanl/bml] Venado optimizations (PR #755) Yes, I think such a function could be good. The dense case is essentially done, even for magma build, we’d just need to add the allocated size to the struct and add a check to make it work. Need to figure out the other matrix types, if they’ll be supported. Meanwhile, what do you think about merging the current function? Get Outlook for iOShttps://aka.ms/o0ukefhttps://urldefense.com/v3/__https://aka.ms/o0ukef__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47KSGeMqG$
________________________________ From: Jean-Luc Fattebert @.> Sent: Monday, January 27, 2025 4:21:48 PM To: lanl/bml @.> Cc: Wall, Michael E @.>; Author @.> Subject: [EXTERNAL] Re: [lanl/bml] Venado optimizations (PR #755 https://urldefense.com/v3/__https://github.com/lanl/bml/pull/755__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47HtHaUCk$) @jeanlucf22 https://urldefense.com/v3/__https://github.com/jeanlucf22__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47HKzc6_0$ commented on this pull request.
________________________________ In src/C-interface/dense/bml_setters_dense.chttps://urldefense.com/v3/#755*discussion_r1931312603;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$: @@ -3,6 +3,22 @@ #include "bml_setters_dense.h" #include "bml_types_dense.h" +#ifdef BML_USE_MAGMA +#include "magma_v2.h" +#endif + +void bml_set_N_dense( I understand now. I was not suggesting to use domain and domain2. I was just saying their allocation may be the main culprit when it comes to allocation time for a dense matrix. Maybe an issue to deal with another time. Another suggestion: having a function resizeNoAlloc(int n) that would just change N if n<=N, otherwise would change N and reallocate memory? Having an extra struct member keeping track of allocated memory size would be good in that case. — Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/#755*discussion_r1931312603;Iw!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8vBVVBy3$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AA67VEIOEH32HB66PF4BJQ32M25QZAVCNFSM6AAAAABVFKYFIOVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDKNZWG42TCNBWGM__;!!Bt8fGhp8LhKGRg!FbQ06GCuZLlZD10ZOodErg20Q_XH3WAFVKL0R0Oc8HqI9Aed_3wlkpMOHRdkO-20l936fvvr2iZGDdyW8iWVuCBR$. You are receiving this because you authored the thread.Message ID: @.> I don't think it is a good idea to merge as is. That's what branches are for if you need it as is right away. — Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https://github.com/lanl/bml/pull/755*issuecomment-2622639216__;Iw!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47BQYtol7$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AA67VEK7MTI6M5EYAESKTYT2NETI7AVCNFSM6AAAAABVFKYFIOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRSGYZTSMRRGY__;!!Bt8fGhp8LhKGRg!Cjdib0KzKOuP85xu8ga136DukYOoPPKVgcctLCsr3CufV8WwR9W6zwe798bqE7qu0gpHiural0miy5b47E1cI_bk$. You are receiving this because you authored the thread.Message ID: @.>

Have the function keep the matrix in a consistent state: memory allocated <= N*ld

mewall · 2025-01-30T16:22:09Z

Have the function keep the matrix in a consistent state: memory allocated <= N*ld

OK here's what I propose

A add a new variable "num_elems_allocated" to the following struct that holds the total number of elements in the dense matrix

bml/src/C-interface/dense/bml_types_dense.h

Line 15 in ef2ee26

struct bml_matrix_dense_t

If you can foresee any consequences for this modification elsewhere in the code, please let me know

Leave bml_set_N_dense() as a function to change N.

Fail using LOG_ERROR()

I will be unable to merge our PROGRESS optimizations into master without merging this into master, so please, if there is anything else that will hold this up, let me know now.

jmohdyusof · 2025-01-30T18:06:57Z

I generally agree with the set/get duality for naming.
So in the case of reusing an allocation, you just set N_new < N_old but retain the same pointer to allocated memory, and keep track of N_original somewhere to make sure N_new <= N_original, otherwise reallocate?

mewall · 2025-01-30T18:17:33Z

I generally agree with the set/get duality for naming. So in the case of reusing an allocation, you just set N_new < N_old but retain the same pointer to allocated memory, and keep track of N_original somewhere to make sure N_new <= N_original, otherwise reallocate?

That's right. The burden is currently on the programmer to ensure the method isn't called when N_new > N_original. Nothing in BML helps with that, AFAIK it would require keeping track of N_original in the struct (the call can happen many times with different values of N_new, so just comparing to the current N doesn't work).

jmohdyusof · 2025-01-30T18:19:59Z

In that case I agree that adding N_original (or N_allocated if you prefer) to the struct is the correct way to go.

mewall · 2025-01-30T18:26:09Z

In that case I agree that adding N_original (or N_allocated if you prefer) to the struct is the correct way to go.

I'm OK with using N_original (linear size) instead of num_elems_allocated (total size of the array). That should work for either the CPU of MAGMA code path.

@jeanlucf22?

jmohdyusof · 2025-01-30T18:29:18Z

I definitely prefer N specification over num_elements.

jeanlucf22 · 2025-01-30T19:34:23Z

In that case I agree that adding N_original (or N_allocated if you prefer) to the struct is the correct way to go.

I'm OK with using N_original (linear size) instead of num_elems_allocated (total size of the array). That should work for either the CPU of MAGMA code path.

@jeanlucf22?

Fine with N_allocated.

mewall · 2025-01-30T19:38:22Z

In that case I agree that adding N_original (or N_allocated if you prefer) to the struct is the correct way to go.

I'm OK with using N_original (linear size) instead of num_elems_allocated (total size of the array). That should work for either the CPU of MAGMA code path.
@jeanlucf22?

Fine with N_allocated.

OK, it sounds like we have a consensus around N_allocated. I'll make the change and push a revision.

jeanlucf22 · 2025-02-03T19:42:06Z

scripts/build_venado_hackathon.sh

+# Make sure all the paths are correct
+
+rm -r build
+#rm -r install_magma_2.7.2


want to clean up this file or remove it?

mewall · 2025-02-03T22:43:41Z

In that case I agree that adding N_original (or N_allocated if you prefer) to the struct is the correct way to go.

I'm OK with using N_original (linear size) instead of num_elems_allocated (total size of the array). That should work for either the CPU of MAGMA code path.
@jeanlucf22?

Fine with N_allocated.

OK, it sounds like we have a consensus around N_allocated. I'll make the change and push a revision.

@jeanlucf22 @jmohdyusof Please check latest commit. N_allocated added to the struct. Check before setting N, do it if new N <= N_allocated, otherwise reallocate the matrix. I'll clean up the lint if it looks OK.

jeanlucf22 · 2025-02-04T18:10:26Z

@mewall If you rebase this branch, it should now pass the tests

Signed-off-by: Nicolas Bock <[email protected]>

Helpful when editing workflow files. Signed-off-by: Nicolas Bock <[email protected]>

Signed-off-by: Nicolas Bock <[email protected]>

mewall added 10 commits January 14, 2025 09:30

Add build script for venado hackathon

e827603

Add preliminary method to set matrix size

096d3fe

Add build scripts for hackathon

eac6576

Add bml_transpose_inplace Fortran subroutine

e7a8fc0

Modify Fortran bml_transpose API to match the C interface

b1dcbb1

Modify tests to use new transpose API

7ded60f

Venado build modifications

87d02dd

Move build scripts to scripts/ dir

d766123

Expose dense matrix pointer using bml_get_ptr_dense

84be6e9

New introspection methods

559bc8e

o Back out bml_get_ptr_dense() from bml_getters o Write fortran wrapper for existing bml_get_data_ptr_dense() method o Write new bml_get_ld_dense() to enable magma matrix pointer use

mewall requested review from nicolasbock, cnegre, suemni, jeanlucf22, tokshgithub and jmohdyusof as code owners January 14, 2025 16:54

suemni approved these changes Jan 14, 2025

View reviewed changes

nicolasbock approved these changes Jan 14, 2025

View reviewed changes

nicolasbock enabled auto-merge January 14, 2025 18:45

jeanlucf22 requested changes Jan 14, 2025

View reviewed changes

scripts/build_bml_cray.sh Outdated Show resolved Hide resolved

src/C-interface/dense/bml_setters_dense.c Outdated Show resolved Hide resolved

Remove LANL-specific build scripts

fc03d39

jeanlucf22 reviewed Jan 27, 2025

View reviewed changes

scripts/build_chicoma_hackathon.sh Outdated Show resolved Hide resolved

Remove an additional LANL specific build script

0fc96d2

Add N_allocated and check in bml_set_N_dense

4918b48

jeanlucf22 reviewed Feb 3, 2025

View reviewed changes

mewall added 2 commits February 3, 2025 12:54

Reallocate if bml_set_N_dense fails test

d6cd30f

Remove LANL-specific build script

1b345ae

jeanlucf22 approved these changes Feb 4, 2025

View reviewed changes

Fix linter issues

2521f0b

nicolasbock and others added 7 commits February 4, 2025 11:13

Fix syntax error in workflow (#758)

c53c933

Signed-off-by: Nicolas Bock <[email protected]>

Update GitHub checkout action (#757)

cec122b

Signed-off-by: Nicolas Bock <[email protected]>

Add yaml extension to dev container (#756)

2d51b8e

Helpful when editing workflow files. Signed-off-by: Nicolas Bock <[email protected]>

Add another VSCode extension to container (#759)

57ab2a0

Signed-off-by: Nicolas Bock <[email protected]>

Update version of artifact action (#761)

30ab5ff

Signed-off-by: Nicolas Bock <[email protected]>

Update compiler for OS X runner (#762)

890b83e

Signed-off-by: Nicolas Bock <[email protected]>

Linter fix

402029c

nicolasbock added this pull request to the merge queue Feb 4, 2025

github-merge-queue bot removed this pull request from the merge queue due to no response for status checks Feb 4, 2025

jeanlucf22 added this pull request to the merge queue Feb 5, 2025

github-merge-queue bot removed this pull request from the merge queue due to no response for status checks Feb 5, 2025

nicolasbock added this pull request to the merge queue Feb 5, 2025

github-merge-queue bot removed this pull request from the merge queue due to no response for status checks Feb 5, 2025

nicolasbock merged commit f386348 into master Feb 5, 2025
31 checks passed

nicolasbock deleted the hackathon branch February 5, 2025 18:37

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Venado optimizations #755

Venado optimizations #755

mewall commented Jan 14, 2025 •

edited

Loading

nicolasbock left a comment

mewall commented Jan 28, 2025 via email

jeanlucf22 commented Jan 29, 2025

jmohdyusof commented Jan 29, 2025

mewall commented Jan 29, 2025 via email

mewall commented Jan 29, 2025 via email

jeanlucf22 commented Jan 30, 2025

mewall commented Jan 30, 2025

jmohdyusof commented Jan 30, 2025

mewall commented Jan 30, 2025

jmohdyusof commented Jan 30, 2025

mewall commented Jan 30, 2025

jmohdyusof commented Jan 30, 2025

jeanlucf22 commented Jan 30, 2025

mewall commented Jan 30, 2025

jeanlucf22 Feb 3, 2025

mewall commented Feb 3, 2025

jeanlucf22 commented Feb 4, 2025

Venado optimizations #755

Venado optimizations #755

Conversation

mewall commented Jan 14, 2025 • edited Loading

nicolasbock left a comment

Choose a reason for hiding this comment

mewall commented Jan 28, 2025 via email

jeanlucf22 commented Jan 29, 2025

jmohdyusof commented Jan 29, 2025

mewall commented Jan 29, 2025 via email

mewall commented Jan 29, 2025 via email

jeanlucf22 commented Jan 30, 2025

mewall commented Jan 30, 2025

jmohdyusof commented Jan 30, 2025

mewall commented Jan 30, 2025

jmohdyusof commented Jan 30, 2025

mewall commented Jan 30, 2025

jmohdyusof commented Jan 30, 2025

jeanlucf22 commented Jan 30, 2025

mewall commented Jan 30, 2025

jeanlucf22 Feb 3, 2025

Choose a reason for hiding this comment

mewall commented Feb 3, 2025

jeanlucf22 commented Feb 4, 2025

mewall commented Jan 14, 2025 •

edited

Loading