On Thu, Jul 28, 2016 at 4:51 PM, Reda Sallahi <[email protected]> wrote: > On 7/28/16, Stefan Hajnoczi <[email protected]> wrote: >> On Wed, Jul 27, 2016 at 3:51 PM, Reda Sallahi <[email protected]> wrote: >>> - qemu_opt_set_number(opts, BLOCK_OPT_SIZE, size, &error_abort); >>> + if (dd.flags & C_SKIP && size < in.bsz * in.offset) { >>> + error_report("%s: cannot skip to specified offset", >>> in.filename); >>> + qemu_opt_set_number(opts, BLOCK_OPT_SIZE, 0, &error_abort); >> >> This looks odd. What is supposed to happen in this case? Do you have >> a test case for it? >> > > We print a warning if we have specified a too big of a value for skip (bigger > than the input size) and set the size to 0 for the output disk image. It's the > same behavior on dd(1). > > In the last iteration of the test case it covers such a case. Though without > checking the output on stderr.
Good that you have a test case. Please also add a comment here explaining that this behavior is compatible with dd and this is a warning rather than an error. Stefan
