[PATCH] firmware: socfpga: Fix a bug in fpgamgr_program_write_buf()

[Andrey Smirnov]

On Tue, Apr 14, 2015 at 11:46 AM, Sascha Hauer <s.hauer at pengutronix.de> wrote:
> Hi Andrey,
>
> On Mon, Apr 13, 2015 at 05:11:50AM -0700, Andrey Smirnov wrote:
>> Fix a bug in fpgamgr_program_write_buf() where .rbf file whose length
>> is not a multiple of 4 would cause an integer overflow which would
>> result in infinite loop.
>>
>> Signed-off-by: Andrey Smirnov <andrew.smirnov at gmail.com>
>> ---
>>  drivers/firmware/socfpga.c | 26 ++++++++++++++++++++++----
>>  1 file changed, 22 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/firmware/socfpga.c b/drivers/firmware/socfpga.c
>> index a5dc607..75fb050 100644
>> --- a/drivers/firmware/socfpga.c
>> +++ b/drivers/firmware/socfpga.c
>> @@ -321,14 +321,32 @@ static int fpgamgr_program_write_buf(struct firmware_handler *fh, const void *bu
>>               size_t size)
>>  {
>>       struct fpgamgr *mgr = container_of(fh, struct fpgamgr, fh);
>> -     const uint32_t *buf32 = buf;
>> +     const uint8_t *buffer = buf;
>> +     uint32_t word;
>> +     size_t chunk_size;
>> +     size_t offset = 0;
>>
>>       /* write to FPGA Manager AXI data */
>>       while (size) {
>
> This code probably looks better when you change this loop to
> while (size >= sizeof(uint32_t)){} followed by a if (size) {}

Yeah, I agree. I'll update it in the next version of the patch.

>
>> -             writel(*buf32, mgr->regs_data);
>> +             chunk_size = min(size, sizeof(uint32_t));
>> +             size -= chunk_size;
>> +
>> +             if (likely(chunk_size == sizeof(uint32_t))) {
>> +                     word = *(uint32_t *)(buffer + offset);
>> +                     offset += sizeof(uint32_t);
>> +             } else {
>> +                     word = buffer[offset++];
>> +                     word <<= 8;
>> +                     chunk_size--;
>> +
>> +                     while (chunk_size--) {
>> +                             word |= buffer[offset++];
>> +                             word <<= 8;
>> +                     }
>
> Isn't this the same as:
>
>                         word = 0;
>                         while (chunk_size--) {
>                                 word |= buffer[offset++];
>                                 word <<= 8;
>                         }
>

That's how I originally wrote it, but then the "unnecessary"
assignment to zero started to bother me for some reason and I
un-rolled one iteration of the loop. I can change it back in the next
version of the patch


>
> Sascha
>
> --
> Pengutronix e.K.                           |                             |
> Industrial Linux Solutions                 | http://www.pengutronix.de/  |
> Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |