blktrace won't help on this issue because .bv_offset isn't recorded. This patch makes sense on >4KB PAGE_SIZE. If your issue happens on ARCH with 4K PAGE_SIZE, maybe you should root cause why it makes a difference on iSer. And it is highly possible there is bug somewhere.
I don't suspect that this is the case. unless all the vec elements happen to have an offset but if that is the case I don't think that this patch would've solved this. It can't really be an iser issue because iser does not control bio splitting/merging... it can only ask the block layer to obay the virt_boundary.
