[PATCH] spi/xilinx: Fix access invalid memory on xilinx_spi_tx

[PATCH] spi/xilinx: Fix access invalid memory on xilinx_spi_tx

[Ricardo Ribalda Delgado]

On 1 and 2 bytes per word, the transfer of the 3 last bytes will access
memory outside rx_ptr.

Although this has not trigger any error on real hardware, we should
better fix this.

Fixes: 24ba5e593f391507 Remove rx_fn and tx_fn pointer
Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
---
 drivers/spi/spi-xilinx.c | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/drivers/spi/spi-xilinx.c b/drivers/spi/spi-xilinx.c
index 2ca55f6..a1b664d 100644
--- a/drivers/spi/spi-xilinx.c
+++ b/drivers/spi/spi-xilinx.c
@@ -97,11 +97,26 @@ struct xilinx_spi {
 
 static void xilinx_spi_tx(struct xilinx_spi *xspi)
 {
+	u32 data = 0;
+
 	if (!xspi->tx_ptr) {
 		xspi->write_fn(0, xspi->regs + XSPI_TXD_OFFSET);
 		return;
 	}
-	xspi->write_fn(*(u32 *)(xspi->tx_ptr), xspi->regs + XSPI_TXD_OFFSET);
+
+	switch (xspi->bytes_per_word) {
+	case 1:
+		data = *(u8 *)(xspi->rx_ptr);
+		break;
+	case 2:
+		data = *(u16 *)(xspi->rx_ptr);
+		break;
+	case 4:
+		data = *(u32 *)(xspi->rx_ptr);
+		break;
+	}
+
+	xspi->write_fn(data, xspi->regs + XSPI_TXD_OFFSET);
 	xspi->tx_ptr += xspi->bytes_per_word;
 }
 
-- 
2.1.4

Re: [PATCH] spi/xilinx: Fix access invalid memory on xilinx_spi_tx

[Geert Uytterhoeven]

Hi Ricardo,

On Fri, Jan 30, 2015 at 7:10 AM, Ricardo Ribalda Delgado
<ricardo.ribalda@gmail.com> wrote:
> On 1 and 2 bytes per word, the transfer of the 3 last bytes will access
> memory outside rx_ptr.
>
> Although this has not trigger any error on real hardware, we should
> better fix this.
>
> Fixes: 24ba5e593f391507 Remove rx_fn and tx_fn pointer
> Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com>
> ---
>  drivers/spi/spi-xilinx.c | 17 ++++++++++++++++-
>  1 file changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/spi/spi-xilinx.c b/drivers/spi/spi-xilinx.c
> index 2ca55f6..a1b664d 100644
> --- a/drivers/spi/spi-xilinx.c
> +++ b/drivers/spi/spi-xilinx.c
> @@ -97,11 +97,26 @@ struct xilinx_spi {
>
>  static void xilinx_spi_tx(struct xilinx_spi *xspi)
>  {
> +       u32 data = 0;
> +
>         if (!xspi->tx_ptr) {
>                 xspi->write_fn(0, xspi->regs + XSPI_TXD_OFFSET);
>                 return;
>         }
> -       xspi->write_fn(*(u32 *)(xspi->tx_ptr), xspi->regs + XSPI_TXD_OFFSET);
> +
> +       switch (xspi->bytes_per_word) {
> +       case 1:
> +               data = *(u8 *)(xspi->rx_ptr);
> +               break;
> +       case 2:
> +               data = *(u16 *)(xspi->rx_ptr);
> +               break;
> +       case 4:
> +               data = *(u32 *)(xspi->rx_ptr);
> +               break;
> +       }
> +
> +       xspi->write_fn(data, xspi->regs + XSPI_TXD_OFFSET);

Is this endian-safe?

>         xspi->tx_ptr += xspi->bytes_per_word;
>  }

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH] spi/xilinx: Fix access invalid memory on xilinx_spi_tx

[Ricardo Ribalda Delgado]

On 30 Jan 2015 11:06, "Geert Uytterhoeven" <geert@linux-m68k.org> wrote:

> Is this endian-safe?

As endianness safe as the original code, but there is also an issue. It is
using rx_ptr.

I am away until  Sunday. So I cannot make a new patch until then. I tried
to make this patch before leaving in the morning and was obviously a
mistake.

Again, Sorry for the mess this is really embarrassing.
Regards
>         xspi->tx_ptr += xspi->bytes_per_word;

> >  }
>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 --
> geert@linux-m68k.org <javascript:;>
>
> In personal conversations with technical people, I call myself a hacker.
> But
> when I'm talking to journalists I just say "programmer" or something like
> that.
>                                 -- Linus Torvalds
>