linux-kernel - [PATCH] genirq: Reenable shared irq debugging in request_*

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 15 Feb 2017 15:43:54 +0100 (CET)
From:   Thomas Gleixner <tglx@...utronix.de>
To:     LKML <linux-kernel@...r.kernel.org>
cc:     Ingo Molnar <mingo@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>
Subject: [PATCH] genirq: Reenable shared irq debugging in request_*_irq()

Commit 6d83f94db95c ("genirq: Disable the SHIRQ_DEBUG call in
request_threaded_irq for now") disabled the SHIRQ_DEBUG facility for the
irq request side. Unfortunately the 'for now' part and the _FIXME add on to
the CONFIG_SHIRQ_DEBUG guard got forgotten.

It's unfortunately not possible to keep the original idea of just invoking
the newly installed handler due to the reasons described in the commit
which disabled it, but having this debug facility available makes a lot of
sense.

The only way to regain that testing is to invoke the real interrupt
handling mechanism, which will also invoke an already registered handler,
but that's not an issue because _all_ interrupt handlers need to be
prepared for handling spurious interrupts.

Signed-off-by: Thomas Gleixner <tglx@...utronix.de>
---
 kernel/irq/manage.c |   10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -1696,23 +1696,17 @@ int request_threaded_irq(unsigned int ir
 		kfree(action);
 	}

-#ifdef CONFIG_DEBUG_SHIRQ_FIXME
+#ifdef CONFIG_DEBUG_SHIRQ
 	if (!retval && (irqflags & IRQF_SHARED)) {
 		/*
 		 * It's a shared IRQ -- the driver ought to be prepared for it
 		 * to happen immediately, so let's make sure....
-		 * We disable the irq to make sure that a 'real' IRQ doesn't
-		 * run in parallel with our fake.
 		 */
 		unsigned long flags;

-		disable_irq(irq);
 		local_irq_save(flags);
-
-		handler(irq, dev_id);
-
+		generic_handle_irq_desc(desc);
 		local_irq_restore(flags);
-		enable_irq(irq);
 	}
 #endif
 	return retval;