Re: [libvirt] [PATCH v2] Edit test capabilities to contain different cell pagesizes.
On 05/23/2018 02:40 PM, Roland Schulz wrote:
Signed-off-by: Roland Schulz <schullzroll(a)gmail.com>
---
src/test/test_driver.c | 18 +++++++++++-------
1 file changed, 11 insertions(+), 7 deletions(-)
In the $SUBJ I'd add the prefix "test driver:" to tell explicitly what
part of code this commit touches.
diff --git a/src/test/test_driver.c b/src/test/test_driver.c
index 467587b19..3fe0c2831 100644
--- a/src/test/test_driver.c
+++ b/src/test/test_driver.c
@@ -322,30 +322,34 @@ testBuildCapabilities(virConnectPtr conn)
if (virCapabilitiesAddHostFeature(caps, "nonpae") < 0)
goto error;
- if (VIR_ALLOC_N(caps->host.pagesSize, 2) < 0)
+ if (VIR_ALLOC_N(caps->host.pagesSize, 4) < 0)
goto error;
caps->host.pagesSize[caps->host.nPagesSize++] = 4;
+ caps->host.pagesSize[caps->host.nPagesSize++] = 8;
caps->host.pagesSize[caps->host.nPagesSize++] = 2048;
+ caps->host.pagesSize[caps->host.nPagesSize++] = 1024 * 1024;
for (i = 0; i < privconn->numCells; i++) {
virCapsHostNUMACellCPUPtr cpu_cells;
virCapsHostNUMACellPageInfoPtr pages;
- size_t nPages;
+ size_t nPages = caps->host.nPagesSize - 1;
if (VIR_ALLOC_N(cpu_cells, privconn->cells[i].numCpus) < 0 ||
- VIR_ALLOC_N(pages, caps->host.nPagesSize) < 0) {
+ VIR_ALLOC_N(pages, caps->host.nPagesSize - 1) < 0) {
or s/caps->host.nPagesSize - 1/nPages/ since they will always have the
same value.
VIR_FREE(cpu_cells);
goto error;
}
-
- nPages = caps->host.nPagesSize;
+ if (i == 1)
+ pages[0].size = caps->host.pagesSize[0];
+ else
+ pages[0].size = caps->host.pagesSize[1];
I'd move this after the memcpy so that pages[] initialization is kept
together with the for() loop below the memcpy(). IOW, memcpy should be
moved before this if() statement.
memcpy(cpu_cells, privconn->cells[i].cpus,
sizeof(*cpu_cells) * privconn->cells[i].numCpus);
- for (j = 0; j < nPages; j++)
- pages[j].size = caps->host.pagesSize[j];
+ for (j = 1; j < nPages; j++)
+ pages[j].size = caps->host.pagesSize[j+1];
Too much spaces ;-) Also, "j + 1" instead of "j+1".
Fixing all the issues, ACKing and pushing.
Michal