On Mon, Nov 28, 2016 at 07:08:42AM -0500, John Ferlan wrote:
>
>
> On 11/25/2016 10:53 AM, Erik Skultety wrote:
> > On Mon, Nov 21, 2016 at 06:35:50PM -0500, John Ferlan wrote:
> >> Rather than have multiple bool values, create a single enum with bits
> >> representing what can be set.
> >>
> >> Signed-off-by: John Ferlan <jferlan@xxxxxxxxxx>
> >> ---
> >> src/qemu/qemu_driver.c | 113 +++++++++++++++++++++++--------------------------
> >> 1 file changed, 54 insertions(+), 59 deletions(-)
> >>
> >> diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
> >> index 87d219f..73b58d0 100644
> >> --- a/src/qemu/qemu_driver.c
> >> +++ b/src/qemu/qemu_driver.c
> >> @@ -17338,34 +17338,38 @@ qemuDomainOpenGraphicsFD(virDomainPtr dom,
> >> return ret;
> >> }
> >>
> >> +typedef enum {
> >> + QEMU_BLOCK_IOTUNE_SET_BYTES = 1 << 0,
> >> + QEMU_BLOCK_IOTUNE_SET_IOPS = 1 << 1,
> >> + QEMU_BLOCK_IOTUNE_SET_BYTES_MAX = 1 << 2,
> >> + QEMU_BLOCK_IOTUNE_SET_IOPS_MAX = 1 << 3,
> >> + QEMU_BLOCK_IOTUNE_SET_SIZE_IOPS = 1 << 4,
> >> + QEMU_BLOCK_IOTUNE_SET_BYTES_MAX_LENGTH = 1 << 5,
> >> + QEMU_BLOCK_IOTUNE_SET_IOPS_MAX_LENGTH = 1 << 6,
> >> +} qemuBlockIoTuneSetFlags;
> >> +
> >>
> >> /* If the user didn't specify bytes limits, inherit previous values;
> >> * likewise if the user didn't specify iops limits. */
> >> static void
> >> qemuDomainSetBlockIoTuneDefaults(virDomainBlockIoTuneInfoPtr newinfo,
> >> virDomainBlockIoTuneInfoPtr oldinfo,
> >> - bool set_bytes,
> >> - bool set_iops,
> >> - bool set_bytes_max,
> >> - bool set_iops_max,
> >> - bool set_size_iops,
> >> - bool set_bytes_max_length,
> >> - bool set_iops_max_length)
> >> + qemuBlockIoTuneSetFlags set_flag)
> >
> > Just a cosmetic "nit", I spent a few seconds looking at the name "set_flag"
> > confusingly (probably 'cause it's Friday). Maybe something like
> > set_map|set_mask|mask|bitmap or something alike would sound better, but then,
> > who am I to judge with my history of 'brilliant' function naming :D.
>
> I could go with set_mask, but mask has a different connotation for some.
> It's not a bitmap and there are those that would say that a bitmap
> should use the Bitmap functions (which would be overkill at this point).
>
> Maybe a 'set_fields_flag' would be clearer? It's a flag to indicate a
the typedef does already have 'flag' in it...how about plain set_fields then?
Erik
> grouping of (usually) 3 elements ({read|write|total}) were set, where
> size_iops and group_name become the exception.
>
> Unless someone else has a brilliant suggestion...
>
> John
> >
> > Patch looks good though, I'll leave it to you.
> >
> > Erik
> >
Attachment:
signature.asc
Description: PGP signature
-- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list
