| 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
| ||
|
Date: Tue, 19 Feb 2013 09:54:16 +0800
From: Frank Li <lznuaa@...il.com>
To: Fabio Estevam <fabio.estevam@...escale.com>
Cc: davem@...emloft.net, Frank.Li@...escale.com, shawn.guo@...aro.org,
marex@...x.de, s.hauer@...gutronix.de, netdev@...r.kernel.org
Subject: Re: [PATCH v2] net: fec: Fix the disabling of RX interrupt
2013/2/19 Fabio Estevam <fabio.estevam@...escale.com>:
> The correct way to disable FEC RX interrupt is to clean only the FEC_ENET_RXF
> bit.
>
> Since commit dc975382d2e (net: fec: add napi support to improve proformance)
> FEC_RX_DISABLED_IMASK is being written to the FEC_IMASK register, which also
> incorrectly sets other bits as per the definitions below:
>
> #define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII)
> #define FEC_RX_DISABLED_IMASK (FEC_DEFAULT_IMASK & (~FEC_ENET_RXF))
>
> This fixes the following kernel crash:
>
> Unable to handle kernel NULL pointer dereference at virtual address 00000002
> pgd = 80004000
> [00000002] *pgd=00000000
> Internal error: Oops: 5 [#1] SMP ARM
> Modules linked in:
> CPU: 0 Not tainted (3.8.0-rc7-next-20130218+ #1192)
> PC is at fec_enet_interrupt+0xd0/0x354
> LR is at fec_enet_interrupt+0xb8/0x354
> pc : [<8034592c>] lr : [<80345914>] psr: 60000193
>
> Signed-off-by: Fabio Estevam <fabio.estevam@...escale.com>
> ---
> Changes since v1:
> - Fix it also inside fec_enet_init and fix the kernel crash.
>
> drivers/net/ethernet/freescale/fec.c | 13 +++++++++----
> 1 file changed, 9 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/net/ethernet/freescale/fec.c b/drivers/net/ethernet/freescale/fec.c
> index 0fe68c4..4eb0201 100644
> --- a/drivers/net/ethernet/freescale/fec.c
> +++ b/drivers/net/ethernet/freescale/fec.c
> @@ -811,6 +811,7 @@ fec_enet_interrupt(int irq, void *dev_id)
> struct fec_enet_private *fep = netdev_priv(ndev);
> uint int_events;
> irqreturn_t ret = IRQ_NONE;
> + int reg;
>
> do {
> int_events = readl(fep->hwp + FEC_IEVENT);
> @@ -821,8 +822,9 @@ fec_enet_interrupt(int irq, void *dev_id)
>
> /* Disable the RX interrupt */
> if (napi_schedule_prep(&fep->napi)) {
> - writel(FEC_RX_DISABLED_IMASK,
> - fep->hwp + FEC_IMASK);
> + reg = readl(fep->hwp + FEC_IMASK);
> + reg &= ~FEC_ENET_RXF;
> + writel(reg, fep->hwp + FEC_IMASK);
This is not correct fix.
Read, OR and write is not atomic. Need lock.
But if use hw_lock, it cause dead lock.
I spend some time to remove such lock.
Can you tell which bit cause crash?
> __napi_schedule(&fep->napi);
> }
> }
> @@ -1598,7 +1600,7 @@ static int fec_enet_init(struct net_device *ndev)
> struct fec_enet_private *fep = netdev_priv(ndev);
> struct bufdesc *cbd_base;
> struct bufdesc *bdp;
> - int i;
> + int i, reg;
>
> /* Allocate memory for buffer descriptors. */
> cbd_base = dma_alloc_coherent(NULL, PAGE_SIZE, &fep->bd_dma,
> @@ -1628,7 +1630,10 @@ static int fec_enet_init(struct net_device *ndev)
> ndev->netdev_ops = &fec_netdev_ops;
> ndev->ethtool_ops = &fec_enet_ethtool_ops;
>
> - writel(FEC_RX_DISABLED_IMASK, fep->hwp + FEC_IMASK);
> + reg = readl(fep->hwp + FEC_IMASK);
> + reg &= ~FEC_ENET_RXF;
> + writel(reg, fep->hwp + FEC_IMASK);
> +
> netif_napi_add(ndev, &fep->napi, fec_enet_rx_napi, FEC_NAPI_WEIGHT);
>
> /* Initialize the receive buffer descriptors. */
> --
> 1.7.9.5
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists