[PATCH 04/10] usb: musb: ux500: harden checks for platform data
In its current state, the ux500-musb driver uses platform data pointers blindly with no prior checking. If no platform data pointer is passed this will Oops the kernel. In this patch we ensure platform data and board data are present prior to using them. Cc: Felipe Balbi Cc: linux-usb@vger.kernel.org Signed-off-by: Lee Jones --- drivers/usb/musb/ux500_dma.c | 11 --- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/usb/musb/ux500_dma.c b/drivers/usb/musb/ux500_dma.c index c75e07a..02b4a6e 100644 --- a/drivers/usb/musb/ux500_dma.c +++ b/drivers/usb/musb/ux500_dma.c @@ -287,7 +287,7 @@ static int ux500_dma_controller_start(struct dma_controller *c) struct musb *musb = controller->private_data; struct device *dev = musb->controller; struct musb_hdrc_platform_data *plat = dev->platform_data; - struct ux500_musb_board_data *data = plat->board_data; + struct ux500_musb_board_data *data; struct dma_channel *dma_channel = NULL; u32 ch_num; u8 dir; @@ -297,14 +297,19 @@ static int ux500_dma_controller_start(struct dma_controller *c) struct ux500_dma_channel *channel_array; dma_cap_mask_t mask; + if (!plat) { + dev_err(musb->controller, "No platform data\n"); + return -EINVAL; + } + data = plat->board_data; dma_cap_zero(mask); dma_cap_set(DMA_SLAVE, mask); /* Prepare the loop for RX channels */ channel_array = controller->rx_channel; - param_array = data->dma_rx_param_array; + param_array = (data) ? data->dma_rx_param_array : NULL; for (dir = 0; dir < 2; dir++) { for (ch_num = 0; @@ -337,7 +342,7 @@ static int ux500_dma_controller_start(struct dma_controller *c) /* Prepare the loop for TX channels */ channel_array = controller->tx_channel; - param_array = data->dma_tx_param_array; + param_array = (data) ? data->dma_tx_param_array : NULL; is_tx = 1; } -- 1.7.10.4 -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 04/10] usb: musb: ux500: harden checks for platform data
Hello. On 04/23/2013 07:03 PM, Lee Jones wrote: In its current state, the ux500-musb driver uses platform data pointers blindly with no prior checking. If no platform data pointer is passed this will Oops the kernel. In this patch we ensure platform data and board data are present prior to using them. Cc: Felipe Balbi Cc: linux-usb@vger.kernel.org Signed-off-by: Lee Jones --- drivers/usb/musb/ux500_dma.c | 11 --- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/usb/musb/ux500_dma.c b/drivers/usb/musb/ux500_dma.c index c75e07a..02b4a6e 100644 --- a/drivers/usb/musb/ux500_dma.c +++ b/drivers/usb/musb/ux500_dma.c @@ -287,7 +287,7 @@ static int ux500_dma_controller_start(struct dma_controller *c) struct musb *musb = controller->private_data; struct device *dev = musb->controller; struct musb_hdrc_platform_data *plat = dev->platform_data; - struct ux500_musb_board_data *data = plat->board_data; + struct ux500_musb_board_data *data; struct dma_channel *dma_channel = NULL; u32 ch_num; u8 dir; @@ -297,14 +297,19 @@ static int ux500_dma_controller_start(struct dma_controller *c) [...] /* Prepare the loop for RX channels */ channel_array = controller->rx_channel; - param_array = data->dma_rx_param_array; + param_array = (data) ? data->dma_rx_param_array : NULL; Why enclose a simple variable in parens? for (dir = 0; dir < 2; dir++) { for (ch_num = 0; @@ -337,7 +342,7 @@ static int ux500_dma_controller_start(struct dma_controller *c) /* Prepare the loop for TX channels */ channel_array = controller->tx_channel; - param_array = data->dma_tx_param_array; + param_array = (data) ? data->dma_tx_param_array : NULL; Again, why? WBR, Sergei -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 04/10] usb: musb: ux500: harden checks for platform data
Hi Sergei, > > struct musb_hdrc_platform_data *plat = dev->platform_data; > >-struct ux500_musb_board_data *data = plat->board_data; > >+struct ux500_musb_board_data *data; > >-param_array = data->dma_rx_param_array; > >+param_array = (data) ? data->dma_rx_param_array : NULL; > >Why enclose a simple variable in parens? Because 'data' is a pointer, so it contains a memory location, but if 'plat->board_data' is NULL, then 'data' will be NULL (essentially memory location 0x). So if we were to read-in to 'struct ux500_musb_board_data *data', by index 'dma_rx_param_array', which I believe is '0' in this case: struct ux500_musb_board_data { void**dma_rx_param_array; void**dma_tx_param_array; bool (*dma_filter)(struct dma_chan *chan, void *filter_param); }; ... then we're saying take the data from this memory location: param_array = *((0x)->(0x0)); Which will cause a kernel Oops, due to the fact that address 0x0 isn't allocated to us, so you get something like: "Unable to handle kernel NULL pointer dereference at virtual address " Hope that helps. Kind regards, Lee -- Lee Jones Linaro ST-Ericsson Landing Team Lead Linaro.org │ Open source software for ARM SoCs Follow Linaro: Facebook | Twitter | Blog -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 04/10] usb: musb: ux500: harden checks for platform data
Hello. On 24-04-2013 10:53, Lee Jones wrote: struct musb_hdrc_platform_data *plat = dev->platform_data; - struct ux500_musb_board_data *data = plat->board_data; + struct ux500_musb_board_data *data; - param_array = data->dma_rx_param_array; + param_array = (data) ? data->dma_rx_param_array : NULL; Why enclose a simple variable in parens? Because 'data' is a pointer, so it contains a memory location, Pointer points to memory location, not contains it. but if 'plat->board_data' is NULL, then 'data' will be NULL (essentially memory location 0x). So what? So if we were to read-in to 'struct ux500_musb_board_data *data', by index 'dma_rx_param_array', which I believe is '0' in this case: struct ux500_musb_board_data { void**dma_rx_param_array; void**dma_tx_param_array; bool (*dma_filter)(struct dma_chan *chan, void *filter_param); }; ... then we're saying take the data from this memory location: param_array = *((0x)->(0x0)); Which will cause a kernel Oops, due to the fact that address 0x0 isn't allocated to us, so you get something like: "Unable to handle kernel NULL pointer dereference at virtual address " We're not dereferencing 'data', so I completely fail to follow you. Hope that helps. Not at all. Kind regards, Lee WBR, Sergei -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 04/10] usb: musb: ux500: harden checks for platform data
On Wed, Apr 24, 2013 at 06:00:28PM +0400, Sergei Shtylyov wrote: > Hello. > > On 24-04-2013 10:53, Lee Jones wrote: > > >>> struct musb_hdrc_platform_data *plat = dev->platform_data; > >>>- struct ux500_musb_board_data *data = plat->board_data; > >>>+ struct ux500_musb_board_data *data; > > >>>- param_array = data->dma_rx_param_array; > >>>+ param_array = (data) ? data->dma_rx_param_array : NULL; > > >>Why enclose a simple variable in parens? > > >Because 'data' is a pointer, so it contains a memory location, heh, I don't think you fully understood Sergei's comment. He's asking why you used: param_array = (data) ? data->dma_rx_param_array : NULL; instead of: param_array = data ? data->dma_rx_param_array : NULL; He's saying, correctly so, that the parens around data is unnecessary. -- balbi signature.asc Description: Digital signature
Re: [PATCH 04/10] usb: musb: ux500: harden checks for platform data
On Wednesday 24 April 2013, Lee Jones wrote: > Yeah, I agree, although does it make a difference? > > Is it evaluated a second time, or does it take up extra cycles by being > enclosed in parentheses? > > Or is this just a coding style thing? Just coding style. I agree you should have no parentheses there, but it does not change the compiled code. Arnd -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html