Re: [PATCH] hw/display/xlnx_dp: fix an out-of-bounds read in xlnx_dp_read

[Qiang Liu]

On Tue, Jul 13, 2021 at 6:24 PM Philippe Mathieu-Daudé <f4bug@amsat.org> wrote:
>
> On 7/13/21 12:20 PM, Philippe Mathieu-Daudé wrote:
> > On 7/13/21 5:14 AM, Qiang Liu wrote:
> >> xlnx_dp_read allows an out-of-bounds read at its default branch because
> >> of an improper index.
> >>
> >> According to
> >> https://www.xilinx.com/html_docs/registers/ug1087/ug1087-zynq-ultrascale-registers.html
> >> (DP Module), registers 0x3A4/0x3A4/0x3AC are allowed.
> >>
> >> DP_INT_MASK  0x000003A4      32      mixed   0xFFFFF03F      Interrupt 
> >> Mask Register for intrN.
> >> DP_INT_EN    0x000003A8      32      mixed   0x00000000      Interrupt 
> >> Enable Register.
> >> DP_INT_DS    0x000003AC      32      mixed   0x00000000      Interrupt 
> >> Disable Register.
> >>
> >> In xlnx_dp_write, when the offset is 0x3A8 and 0x3AC, the virtual device
> >> will write s->core_registers[0x3A4
> >>>> 2]. That is to say, the maxize of s->core_registers could be ((0x3A4
> >>>> 2) + 1). However, the current size of s->core_registers is (0x3AF >>
> >>>> 2), that is ((0x3A4 >> 2) + 2), which is out of the range.
> >> In xlxn_dp_read, the access to offset 0x3A8 or 0x3AC will be directed to
> >> the offset 0x3A8 (incorrect functionality) or 0x3AC (out-of-bounds read)
> >> rather than 0x3A4.
> >>
> >> This patch adjusts the size of s->core_registers and enforces the read
> >> access to offset 0x3A* and 0x3AC to 0x3A4. BTW, because the size of this
> >> MMIO region is 0x3AF, this patch also removes the assertion in
> >> xlnx_dp_write.
> >>
> >> Fixes: 58ac482a66de ("introduce xlnx-dp")
> >> Signed-off-by: Qiang Liu <cyruscyliu@gmail.com>
> >> ---
> >>  hw/display/xlnx_dp.c         | 7 ++++---
> >>  include/hw/display/xlnx_dp.h | 2 +-
> >>  2 files changed, 5 insertions(+), 4 deletions(-)
> >
> > Can you provide a qtest reproducer please?
>
> See examples in tests/qtest/fuzz*test.c
Yeah. I can add the qtest reproducer.

> >> diff --git a/hw/display/xlnx_dp.c b/hw/display/xlnx_dp.c
> >> index 7bcbb13..8903181 100644
> >> --- a/hw/display/xlnx_dp.c
> >> +++ b/hw/display/xlnx_dp.c
> >> @@ -713,8 +713,10 @@ static uint64_t xlnx_dp_read(void *opaque, hwaddr 
> >> offset, unsigned size)
> >>          ret = 0;
> >>          break;
> >>      default:
> >> -        assert(offset <= (0x3AC >> 2));
> >> -        ret = s->core_registers[offset];
> >> +        if (offset == (0x3A8 >> 2) || offset == (0x3AC >> 2))
> >> +            ret = s->core_registers[DP_INT_MASK];
> >> +        else
> >
> > Invalid code style.
Will refine it.

> >> +            ret = s->core_registers[offset];
> >>          break;
> >>      }
> >>
> >> @@ -876,7 +878,6 @@ static void xlnx_dp_write(void *opaque, hwaddr offset, 
> >> uint64_t value,
> >>          xlnx_dp_update_irq(s);
> >>          break;
> >>      default:
> >> -        assert(offset <= (0x504C >> 2));
> >>          s->core_registers[offset] = value;
> >>          break;
> >>      }
> >> diff --git a/include/hw/display/xlnx_dp.h b/include/hw/display/xlnx_dp.h
> >> index e85e428..99a6d47 100644
> >> --- a/include/hw/display/xlnx_dp.h
> >> +++ b/include/hw/display/xlnx_dp.h
> >> @@ -39,7 +39,7 @@
> >>  #define AUD_CHBUF_MAX_DEPTH                 (32 * KiB)
> >>  #define MAX_QEMU_BUFFER_SIZE                (4 * KiB)
> >>
> >> -#define DP_CORE_REG_ARRAY_SIZE              (0x3AF >> 2)
> >> +#define DP_CORE_REG_ARRAY_SIZE              (0x3A8 >> 2)
> >
> > NAck: this breaks migration.
I will not modify this.

> >>  #define DP_AVBUF_REG_ARRAY_SIZE             (0x238 >> 2)
> >>  #define DP_VBLEND_REG_ARRAY_SIZE            (0x1DF >> 2)
> >>  #define DP_AUDIO_REG_ARRAY_SIZE             (0x50 >> 2)

Thank you for the review. I will resend the patch soon.

Best,
Qiang