redonkable/remarkable-linux
redonkable/remarkable-linux
1
0
Fork 0
Code Releases Activity

xen-netback: fix race between napi_complete() and interrupt handler

Browse source 
When the NAPI budget was not all used, xenvif_poll() would call
napi_complete() /after/ enabling the interrupt.  This resulted in a
race between the napi_complete() and the napi_schedule() in the
interrupt handler.  The use of local_irq_save/restore() avoided by
race iff the handler is running on the same CPU but not if it was
running on a different CPU.

Fix this properly by calling napi_complete() before reenabling
interrupts (in the xenvif_napi_schedule_or_enable_irq() call).

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
David Vrabel 2014-05-16 12:26:04 +01:00 committed by David S. Miller
parent 202630b445
commit 0d08fceb2e
3 changed files with 6 additions and 30 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
drivers/net/xen-netback/common.h
View file

@ -226,7 +226,7 @@ int xenvif_map_frontend_rings(struct xenvif *vif,
grant_ref_t rx_ring_ref);
/* Check for SKBs from frontend and schedule backend processing */
void xenvif_check_rx_xenvif(struct xenvif *vif);
void xenvif_napi_schedule_or_enable_events(struct xenvif *vif);
/* Prevent the device from generating any further traffic. */
void xenvif_carrier_off(struct xenvif *vif);

30
drivers/net/xen-netback/interface.c
View file

@ -75,32 +75,8 @@ static int xenvif_poll(struct napi_struct *napi, int budget)
work_done = xenvif_tx_action(vif, budget);
if (work_done < budget) {
int more_to_do = 0;
unsigned long flags;
/* It is necessary to disable IRQ before calling
* RING_HAS_UNCONSUMED_REQUESTS. Otherwise we might
* lose event from the frontend.
*
* Consider:
* RING_HAS_UNCONSUMED_REQUESTS
* <frontend generates event to trigger napi_schedule>
* __napi_complete
*
* This handler is still in scheduled state so the
* event has no effect at all. After __napi_complete
* this handler is descheduled and cannot get
* scheduled again. We lose event in this case and the ring
* will be completely stalled.
*/
local_irq_save(flags);
RING_FINAL_CHECK_FOR_REQUESTS(&vif->tx, more_to_do);
if (!more_to_do)
__napi_complete(napi);
local_irq_restore(flags);
napi_complete(napi);
xenvif_napi_schedule_or_enable_events(vif);
}
return work_done;
@ -194,7 +170,7 @@ static void xenvif_up(struct xenvif *vif)
enable_irq(vif->tx_irq);
if (vif->tx_irq != vif->rx_irq)
enable_irq(vif->rx_irq);
xenvif_check_rx_xenvif(vif);
xenvif_napi_schedule_or_enable_events(vif);
}
static void xenvif_down(struct xenvif *vif)

4
drivers/net/xen-netback/netback.c
View file

@ -716,7 +716,7 @@ done:
notify_remote_via_irq(vif->rx_irq);
}
void xenvif_check_rx_xenvif(struct xenvif *vif)
void xenvif_napi_schedule_or_enable_events(struct xenvif *vif)
{
int more_to_do;
@ -750,7 +750,7 @@ static void tx_credit_callback(unsigned long data)
{
struct xenvif *vif = (struct xenvif *)data;
tx_add_credit(vif);
xenvif_check_rx_xenvif(vif);
xenvif_napi_schedule_or_enable_events(vif);
}
static void xenvif_tx_err(struct xenvif *vif,