AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

SWDEV-272089 - Shorten memory test time

Browse Source 
memory/hipMalloc_MultiThreaded_MultiGpu costs too much time to finish.
    1 GPU: about 1000s, 2 GPUs: about 2200s
But Jenkin build need quick return and ctest will kill test that last
1500+s. So we need shorten the test time.

Change-Id: I3db27a962808566ea135e3cbdefa66d373e2d369


[ROCm/clr commit: 657a515e60]
This commit is contained in:
Tao Sang
2021-02-22 18:00:24 -05:00
parent fc6ff00574
commit ae1db33145
1 changed files with 6 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/clr/hipamd/tests/src/runtimeApi/memory/hipMallocConcurrency.cpp
+6 -2
View File
@@ -85,11 +85,15 @@ Testcase Scenarios :
/* Buffer size for alloc/free cycles */
#define BUFF_SIZE_AF (5*1024*1024)
/* You may change it for individual test.
* But default 100 is for quick return in Jenkin Build */
#define NUM_DIV 100
/* Max alloc/free iterations for smaller chunks */
#define MAX_ALLOCFREE_SC (5000000)
#define MAX_ALLOCFREE_SC (5000000/NUM_DIV)
/* Max alloc and pool iterations (TBD) */
#define MAX_ALLOCPOOL_ITER (2000000)
#define MAX_ALLOCPOOL_ITER (2000000/NUM_DIV)
/**
* Validates data consitency on supplied gpu