Re: [Qemu-devel] [PATCH 1/9] exec: add endian specific phys ld/st functions

[Paolo Bonzini <pbonzini@redhat.com>]

On 07/06/2011 03:03 PM, Hannes Reinecke wrote:
>
> uint32_t ldub_phys(target_phys_addr_t addr);
> uint32_t lduw_phys(target_phys_addr_t addr);
>
> Hmm? ldub is supposed to read an 'unsigned byte' (uint8_t),
> and lduw is supposed to read an 'unsigned word' (uint16_t).
>
> Why does it return an uint32_t?

I don't know if this is the reason, but uint{8,16}_t are promoted to a 
signed int.  So when you do

   (uint64_t) (ldub_phys (addr) << 24)

you'd get a sign extension in bits 32-63.  Admittedly a bit contrived, 
but it can happen and QEMU is full of such bugs:

     case 4:
         lba = (uint64_t) buf[9] | ((uint64_t) buf[8] << 8) |
               ((uint64_t) buf[7] << 16) | ((uint64_t) buf[6] << 24) |
               ((uint64_t) buf[5] << 32) | ((uint64_t) buf[4] << 40) |
               ((uint64_t) buf[3] << 48) | ((uint64_t) buf[2] << 56);
         break;

(found by Coverity).

This was the reason for my series at 
http://permalink.gmane.org/gmane.comp.emulators.qemu/105336 (which you 
reminded me to ping---thanks!)

Paolo