linux-kernel - [PATCH] change lock model in aio_put

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, 6 Jun 2012 22:46:52 +0000
From:	mihailov ivan <mihailov.iaa@...il.com>
To:	linux-kernel@...r.kernel.org
Subject: [PATCH] change lock model in aio_put_req

Why used spin_lock/unlock_irq instead of
spin_lock_irqsave/spin_unlock_irqrestore?
__aio_put_req it's interrupt safe call but why aio_put_req not?

simply if I'll do some kind of this code:

spin_lock_irqsave(&lock, flags);
--some functions 1--
aio_put_req(aiocb);
--some functions 2--
spin_unlock_irqrestore(&lock, flags);

On aio_put_req we unlock all interrupts...
And after that 'some functions 2' unsafe now.

Or this call work as designed?
If not I have smally patchset for that.


--- /usr/src/linux/fs/aio.c	2011-07-22 02:17:23.000000000 +0000
+++ aio.c	2012-06-07 01:12:33.737299121 +0000
@@ -581,10 +581,10 @@
 int aio_put_req(struct kiocb *req)
 {
 	struct kioctx *ctx = req->ki_ctx;
-	int ret;
-	spin_lock_irq(&ctx->ctx_lock);
+	int ret, flags;
+	spin_lock_irqsave(&ctx->ctx_lock, flags);
 	ret = __aio_put_req(ctx, req);
-	spin_unlock_irq(&ctx->ctx_lock);
+	spin_unlock_irqrestore(&ctx->ctx_lock, flags);
 	return ret;
 }
 EXPORT_SYMBOL(aio_put_req);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/