On Wed, Oct 09, 2019 at 01:20:17PM +0530, Ritesh Harjani wrote: > On 10/9/19 12:37 PM, Christoph Hellwig wrote: > > On Wed, Oct 09, 2019 at 11:32:54AM +0530, Ritesh Harjani wrote: > > > We can also get rid of "first_block" argument here. > > > > That would just duplicate filling it out in all callers, so why? > > > > What I meant is "first_block" is same as map->m_lblk. > (unless ext4_map_blocks can change map->m_lblk, which AFAICT, it should > not). > So why pass an extra argument when we are already passing 'map' > structure. > So we can fill iomap->offset using map->m_lblk in ext4_set_iomap() > function. What you're saying makes sense Ritesh and I will update it as such. I believe what Christoph was actually referring to was what I explained to Jan to the email that I just sent out. This was around the possible code duplication and having some iomap value setting related logic outside ext4_set_iomap(), when in fact there's no real reason why it can't be inside. --<M>--
