[PATCH] mptctl.c memory leak on failure

[PATCH] mptctl.c memory leak on failure

[Arnaldo Carvalho de Melo]

Hi,

	Please consider applying this patch. And I got confused by the
kmalloc and the comment: the buffer is used for DMA but the kmalloc doesn't
has GFP_DMA, maybe I'm missing something here, its about time for me to
sleep 8) but if it doesn't needs DMA you should consider using vmalloc and
you'll get rid of the 32 pages limitation, like I did for wanrouter that
used kmalloc but didn't needed DMA, so I've changed it to vmalloc as per
Alan's cluebating^Wsuggestion at the time 8)

- Arnaldo

--- linux-2.4.0-ac3/drivers/message/fusion/mptctl.c	Sat Jan  6 23:13:11 2001
+++ linux-2.4.0-ac3.acme/drivers/message/fusion/mptctl.c	Sun Jan  7 01:02:52 2001
@@ -290,6 +290,7 @@
 	dma_addr_t		 fwbuf_dma;
 	FWDownloadTCSGE_t	*fwVoodoo;
 	SGEAllUnion_t		*fwSgl;
+	int ret;
 
 	dprintk((KERN_INFO "mptctl_do_fwdl called. mptctl_id = %xh.\n", mptctl_id)); //tc
 
@@ -309,20 +310,22 @@
 	}
 //dprintk((KERN_INFO "DbG: fwbuf       = %p\n", fwbuf));
 
+	ret = -EFAULT;
 	if (copy_from_user(fwbuf, ufwbuf, fwlen) < 0) {
 		printk(KERN_ERR "%s@%d::_ioctl_fwdl - "
 				"Unable to copy f/w buffer @ %p\n",
 				__FILE__, __LINE__, (void*)ufwbuf);
-		return -EFAULT;
+		goto out;
 	}
 
 	/* Perform Firmware download.
 	 */
 
+	ret = -ENXIO; /* (-6) No such device or address */
 	if ((ioc = mpt_verify_adapter(ioc, &iocp)) < 0) {
 		printk("%s@%d::_ioctl_fwdl - ioc%d not found!\n",
 				__FILE__, __LINE__, ioc);
-		return -ENXIO;			  /* (-6) No such device or address */
+		goto out;
 	}
 
 	mf = MptGetMsgFrame(mptctl_id, ioc);
@@ -373,8 +376,9 @@
 					   "In ReplyMsg loop - iteration %d\n",
 					   foo)); //tc
 			}
+			ret = -ETIME;
 			if (++foo > 60000000)
-				return -ETIME;
+				goto out;
 			mb();
 			schedule();
 			barrier();
@@ -402,6 +406,9 @@
 		printk(KERN_WARNING MYNAM ": (bad VooDoo):\n");
 		return -ENOMSG;
 	}
+	return 0;
+out:	kfree(fwbuf);
+	return ret;
 }
 
 /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: [PATCH] mptctl.c memory leak on failure

[Alan Cox]

> kmalloc and the comment: the buffer is used for DMA but the kmalloc doesn't
> has GFP_DMA, maybe I'm missing something here, its about time for me to

It should be kmalloc (or 2.4 wise pci_alloc_* I guess eventually). Its driven
by 32bit busmaster DMA. Its non ISA so it doesnt need GFP_DMA
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: [PATCH] mptctl.c memory leak on failure

[Arnaldo Carvalho de Melo]

Em Sun, Jan 07, 2001 at 01:37:15PM +0000, Alan Cox escreveu:
> > kmalloc and the comment: the buffer is used for DMA but the kmalloc doesn't
> > has GFP_DMA, maybe I'm missing something here, its about time for me to
> 
> It should be kmalloc (or 2.4 wise pci_alloc_* I guess eventually). Its driven
> by 32bit busmaster DMA. Its non ISA so it doesnt need GFP_DMA

yap, I should have have some sleep before suggesting that 8) Now back
looking at the other files, I've found another not so grave potential
memory leak, patch in some minutes

- Arnaldo
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/