PCI: hv: Use tasklet_disable_in_atomic()
authorSebastian Andrzej Siewior <bigeasy@linutronix.de>
Tue, 9 Mar 2021 08:42:15 +0000 (09:42 +0100)
committerThomas Gleixner <tglx@linutronix.de>
Wed, 17 Mar 2021 15:34:03 +0000 (16:34 +0100)
The hv_compose_msi_msg() callback in irq_chip::irq_compose_msi_msg is
invoked via irq_chip_compose_msi_msg(), which itself is always invoked from
atomic contexts from the guts of the interrupt core code.

There is no way to change this w/o rewriting the whole driver, so use
tasklet_disable_in_atomic() which allows to make tasklet_disable()
sleepable once the remaining atomic users are addressed.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Wei Liu <wei.liu@kernel.org>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20210309084242.516519290@linutronix.de
drivers/pci/controller/pci-hyperv.c

index 27a17a1..a313708 100644 (file)
@@ -1458,7 +1458,7 @@ static void hv_compose_msi_msg(struct irq_data *data, struct msi_msg *msg)
         * Prevents hv_pci_onchannelcallback() from running concurrently
         * in the tasklet.
         */
-       tasklet_disable(&channel->callback_event);
+       tasklet_disable_in_atomic(&channel->callback_event);
 
        /*
         * Since this function is called with IRQ locks held, can't