On Wed, Aug 24, 2022 at 02:44:06PM +0200, Greg KH wrote:
On Wed, Aug 24, 2022 at 06:00:07PM +0530, Manivannan Sadhasivam wrote:
"pci_endpoint_test" driver now returns 0 for success and negative error code for failure. So adapt to the change by reporting FAILURE if the return value is < 0, and SUCCESS otherwise.
Cc: stable@vger.kernel.org #5.10 Fixes: 3f2ed8134834 ("tools: PCI: Add a userspace tool to test PCI endpoint") Signed-off-by: Manivannan Sadhasivam manivannan.sadhasivam@linaro.org
tools/pci/pcitest.c | 41 +++++++++++++++++++++-------------------- 1 file changed, 21 insertions(+), 20 deletions(-)
diff --git a/tools/pci/pcitest.c b/tools/pci/pcitest.c index 441b54234635..a4e5b17cc3b5 100644 --- a/tools/pci/pcitest.c +++ b/tools/pci/pcitest.c @@ -18,7 +18,6 @@ #define BILLION 1E9 -static char *result[] = { "NOT OKAY", "OKAY" }; static char *irq[] = { "LEGACY", "MSI", "MSI-X" }; struct pci_test { @@ -54,9 +53,9 @@ static int run_test(struct pci_test *test) ret = ioctl(fd, PCITEST_BAR, test->barnum); fprintf(stdout, "BAR%d:\t\t", test->barnum); if (ret < 0)
fprintf(stdout, "TEST FAILED\n");
fprintf(stdout, "FAILED\n");
fprintf(stdout, "%s\n", result[ret]);
fprintf(stdout, "SUCCESS\n");Is this following the kernel TAP output rules? If not, why not? If so, say that you are fixing that issue up in the changelog text.
Sorry to revive this two months old thread. Adapting to TAP output rules requires this test to be moved to KUnit which is strictly not necessary and can be done later.
Moreover, I do not have the hardware to run this testcase and I don't feel comfortable moving this to KUnit without doing functional testing.
So for now, I will fix the return value of IOCTLs which is the real motive behind this series.
Thanks, Mani
thanks,
greg k-h