On 6/25/19 6:38 PM, Damien Le Moal wrote: > On 2019/06/26 1:06, Chaitanya Kulkarni wrote: >> On 06/25/2019 08:59 AM, Bart Van Assche wrote: >>> On 6/24/19 8:24 PM, Chaitanya Kulkarni wrote: >>>> nit:- Can we use is_vmalloc_addr() call directly so that >>>> "if (is_vmalloc)" -> "if (is_vmalloc_addr(data))" and remove is_vmalloc >>>> variable. >>> That would change a single call of is_vmalloc_addr() into multiple? >> >> Well is_vmalloc_addr() it is an in-line helper with address comparison. >> >> is it too expensive to have such a comparison in the loop ? > > Probably not, but I do not see the point in calling it for every page either > since the cost of the additional bool on the stack is likely also very low. > Okay.
