Hi drew,
On 11/29/23 19:43, Andrew Jones wrote:
On Wed, Nov 29, 2023 at 06:14:20PM +0800, Shaoqin Huang wrote:
Hi drew,
On 11/29/23 17:27, Andrew Jones wrote:
On Tue, Nov 28, 2023 at 10:21:23PM -0500, Shaoqin Huang wrote:
Currently running tests on EFI in parallel can cause part of tests to
fail, this is because arm/efi/run script use the EFI_CASE to create the
subdir under the efi-tests, and the EFI_CASE is the filename of the
test, when running tests in parallel, the multiple tests exist in the
same filename will execute at the same time, which will use the same
directory and write the test specific things into it, this cause
chaotic and make some tests fail.
How do they fail? iiuc, we're switching from having one of each unique
binary on the efi partition to multiple redundant binaries, since we
copy the binary for each test, even when it's the same as for other
tests. It seems like we should be able to keep single unique binaries
and resolve the parallel execution failure by just checking for existence
of the binaries or only creating test-specific data directories or
something.
The problem comes from the arm/efi/run script, as you can see. If we
parallel running multiple tests on efi, for example, running the pmu-sw-incr
and pmu-chained-counters and other pmu tests at the same time, the EFI_CASE
will be pmu. So they will write their $cmd_args to the
$EFI/TEST/pmu/startup.nsh at the same time, which will corrupt the
startup.nsh file.
cp "$EFI_SRC/$EFI_CASE.efi" "$EFI_TEST/$EFI_CASE/"
echo "@echo -off" > "$EFI_TEST/$EFI_CASE/startup.nsh"
...
echo "$EFI_CASE.efi" "${cmd_args[@]}" >> "$EFI_TEST/$EFI_CASE/startup.nsh"
And I can get the log which outputs:
* pmu-sw-incr.log:
- ABORT: pmu: Unknown sub-test 'pmu-mem-acce'
* pmu-chained-counters.log
- ABORT: pmu: Unknown sub-test 'pmu-mem-access-reliab'
And the efi-tests/pmu/startup.nsh:
@echo -off
setvar fdtfile -guid 97ef3e03-7329-4a6a-b9ba-6c1fdcc5f823 -rt =L"dtb"
pmu.efi pmu-mem-access-reliability
setvar fdtfile -guid 97ef3e03-7329-4a6a-b9ba-6c1fdcc5f823 -rt =L"dtb"
pmu.efi pmu-chained-sw-incr
Thus when running parallel, some of them will fail. So I create different
sub-dir in the efi-tests for each small test.
Ok, I was guessing it was something like that. Maybe we should create a
"bin" type of named directory for all the binaries and then create a
separate subdir for each test and its startup.nsh, rather than copying
the binaries multiple times.
I can put this kind of improvement into another patch. Currently I want
to use the safety way to fix this problem, create different directory
and copy the .efi for each of them will not cause any interfere for each
of the test. We can tolerate the redundant, it doesn't cost many space.
Thanks,
Shaoqin
: "${EFI_CASE:=$(basename $1 .efi)}"
: "${EFI_VAR_GUID:=97ef3e03-7329-4a6a-b9ba-6c1fdcc5f823}"
@@ -56,20 +58,20 @@ if [ "$EFI_CASE" = "_NO_FILE_4Uhere_" ]; then
EFI_CASE=dummy
fi
-: "${EFI_CASE_DIR:="$EFI_TEST/$EFI_CASE"}"
+: "${EFI_CASE_DIR:="$EFI_TEST/$EFI_TESTNAME"}"
mkdir -p "$EFI_CASE_DIR"
-cp "$EFI_SRC/$EFI_CASE.efi" "$EFI_TEST/$EFI_CASE/"
-echo "@echo -off" > "$EFI_TEST/$EFI_CASE/startup.nsh"
+cp "$EFI_SRC/$EFI_CASE.efi" "$EFI_CASE_DIR/"
+echo "@echo -off" > "$EFI_CASE_DIR/startup.nsh"
Unrelated change, should be a separate patch.
Ok, Will separate it.
Actually, disregard my comment. I was too hasty here and thought the
'@echo -off' was getting added, but now I see only the path was getting
updated. It's correct to make this change here in this patch.
Thanks,
drew
--
Shaoqin
