On 10/21/2021 4:28 AM, Christoph Hellwig wrote:
>> + if (flags & DAXDEV_F_RECOVERY) {
>> + lead_off = (unsigned long)addr & ~PAGE_MASK;
>> + len = PFN_PHYS(PFN_UP(lead_off + bytes));
>> + if (is_bad_pmem(&pmem->bb, PFN_PHYS(pgoff) / 512, len)) {
>> + if (lead_off || !(PAGE_ALIGNED(bytes))) {
>> + dev_warn(dev, "Found poison, but addr(%p) and/or bytes(%#lx) not page aligned\n",
>> + addr, bytes);
>> + return (size_t) -EIO;
>> + }
>> + pmem_off = PFN_PHYS(pgoff) + pmem->data_offset;
>> + if (pmem_clear_poison(pmem, pmem_off, bytes) !=
>> + BLK_STS_OK)
>> + return (size_t) -EIO;
>> + }
>
> Shouldn't this just go down in a separe ->clear_poison operation
> to make the whole thing a little easier to follow?
>
Do you mean to lift or refactor the above to a helper function so as
to improve the readability of the code? I can do that, just to confirm.
On the same note, would you prefer to refactor the read path as well?
thanks!
-jane
