There are two duplicated calls to cleanup_domain_irq_pirq in
unmap_domain_pirq.
The first one in the for loop will be called with exactly the same
arguments as the call placed closer to the loop start.
The second one will only be executed when desc != NULL, and that's
already covered by the first call in the for loop above, as any
attempt to unmap a multi vector MSI range will have nr != 1 and thus
always exit the loop with desc == NULL.
Note that those calls are harmless, but make the code harder to read.
No functional change intended.
Signed-off-by: Roger Pau Monné <roger....@citrix.com>
---
The logic used in the loop seems extremely complex to follow IMO,
there are several breaks for exiting the loop, and the index (i) is
also updated in different places.
---
xen/arch/x86/irq.c | 6 ------
1 file changed, 6 deletions(-)
diff --git a/xen/arch/x86/irq.c b/xen/arch/x86/irq.c
index 032fe82167..49849bd7d3 100644
--- a/xen/arch/x86/irq.c
+++ b/xen/arch/x86/irq.c
@@ -2383,9 +2383,6 @@ int unmap_domain_pirq(struct domain *d, int pirq)
spin_unlock_irqrestore(&desc->lock, flags);
- if ( !forced_unbind )
- cleanup_domain_irq_pirq(d, irq, info);
-
rc = irq_deny_access(d, irq);
if ( rc )
{
@@ -2419,9 +2416,6 @@ int unmap_domain_pirq(struct domain *d, int pirq)
{
spin_unlock_irqrestore(&desc->lock, flags);
- if ( !forced_unbind )
- cleanup_domain_irq_pirq(d, irq, info);
-
rc = irq_deny_access(d, irq);
if ( rc )
{
--
2.29.2
