summaryrefslogtreecommitdiffstats
path: root/dom/webgpu/tests/cts/checkout/src/webgpu/api/operation/memory_allocation/README.txt
diff options
context:
space:
mode:
Diffstat (limited to 'dom/webgpu/tests/cts/checkout/src/webgpu/api/operation/memory_allocation/README.txt')
-rw-r--r--dom/webgpu/tests/cts/checkout/src/webgpu/api/operation/memory_allocation/README.txt7
1 files changed, 7 insertions, 0 deletions
diff --git a/dom/webgpu/tests/cts/checkout/src/webgpu/api/operation/memory_allocation/README.txt b/dom/webgpu/tests/cts/checkout/src/webgpu/api/operation/memory_allocation/README.txt
new file mode 100644
index 0000000000..a8a8eb1d68
--- /dev/null
+++ b/dom/webgpu/tests/cts/checkout/src/webgpu/api/operation/memory_allocation/README.txt
@@ -0,0 +1,7 @@
+Try to stress memory allocators in the implementation and driver.
+
+TODO: plan and implement
+- Tests which (pseudo-randomly?) allocate a bunch of memory and then assert things about the memory
+ (it's not aliased, it's valid to read and write in various ways, accesses read/write the correct data)
+ - Possibly also with OOB accesses/robust buffer access?
+- Tests which are targeted against particular known implementation details