2011/4/4 Daniel Veillard <veillard(a)redhat.com>:
On Sun, Apr 03, 2011 at 02:19:14PM +0200, Matthias Bolte wrote:
> ---
> src/vmx/vmx.c | 10 +++++-----
> 1 files changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/src/vmx/vmx.c b/src/vmx/vmx.c
> index 9f4d5fb..9a482ef 100644
> --- a/src/vmx/vmx.c
> +++ b/src/vmx/vmx.c
> @@ -730,7 +730,7 @@ virVMXGetConfigLong(virConfPtr conf, const char *name, long long
*number,
> }
> }
>
> - if (STREQ(value->str, "unlimited")) {
> + if (STRCASEEQ(value->str, "unlimited")) {
> *number = -1;
> } else if (virStrToLong_ll(value->str, NULL, 10, number) < 0) {
> VMX_ERROR(VIR_ERR_INTERNAL_ERROR,
> @@ -1385,7 +1385,7 @@ virVMXParseConfig(virVMXContext *ctx, virCapsPtr caps, const
char *vmx)
> goto cleanup;
> }
>
> - if (sched_cpu_affinity != NULL && STRNEQ(sched_cpu_affinity,
"all")) {
> + if (sched_cpu_affinity != NULL && STRCASENEQ(sched_cpu_affinity,
"all")) {
> const char *current = sched_cpu_affinity;
> int number, count = 0;
>
> @@ -2107,7 +2107,7 @@ virVMXParseDisk(virVMXContext *ctx, virCapsPtr caps, virConfPtr
conf,
> goto cleanup;
> }
> } else if (virFileHasSuffix(fileName, ".iso") ||
> - STREQ(deviceType, "atapi-cdrom")) {
> + STRCASEEQ(deviceType, "atapi-cdrom")) {
> /*
> * This function was called in order to parse a harddisk device,
> * but .iso files and 'atapi-cdrom' devices are for CDROM
devices
> @@ -2146,7 +2146,7 @@ virVMXParseDisk(virVMXContext *ctx, virCapsPtr caps, virConfPtr
conf,
> * handle it.
> */
> goto ignore;
> - } else if (STREQ(deviceType, "atapi-cdrom")) {
> + } else if (STRCASEEQ(deviceType, "atapi-cdrom")) {
> (*def)->type = VIR_DOMAIN_DISK_TYPE_BLOCK;
> (*def)->src = fileName;
>
> @@ -2174,7 +2174,7 @@ virVMXParseDisk(virVMXContext *ctx, virCapsPtr caps, virConfPtr
conf,
> if ((*def)->src == NULL) {
> goto cleanup;
> }
> - } else if (fileType != NULL && STREQ(fileType, "device"))
{
> + } else if (fileType != NULL && STRCASEEQ(fileType,
"device")) {
> (*def)->type = VIR_DOMAIN_DISK_TYPE_BLOCK;
> (*def)->src = fileName;
Sounds fine to me but would you categorize this as bug fix (for 0.9.0)
or rather an improvement and wait. I guess it depends how you ended up
crossing this, and I can't guess it :-)
Daniel
There was a bug report a while ago about the VMX parser being
case-sensitive, but the .vmx files actually being case-insensitive.
The initial fix was incomplete and I just came across this remaining
case-sensitive compares while adding persistent CPU shares.
As 0.9.0 is out of the door now, I pushed it.
Matthias