Re: fadvise+sync_file_range Vs O_DIRECT

On 4/4/07, Phillip Susi <[EMAIL PROTECTED]> wrote: Ameya Mitragotri wrote: > Hello, > I am doing some benchmarking on writes to the hdd from a simple > application (attached). > Im benchmarking the writes using the following > - write with O_DIRECT > - write with fadvise

fadvise+sync_file_range Vs O_DIRECT

Hello, I am doing some benchmarking on writes to the hdd from a simple application (attached). Im benchmarking the writes using the following - write with O_DIRECT - write with fadvise + O_SYNC - write with fadvise + sync_file_range Ive tested the writes in chunks of 4/8/16 KB and 1MB The result

in-kernel data transfer between devices

Hello, I'm trying to write data from device memory (VRAM in this case) to the HDD with a kernel module. I want to avoid using the cache so i tried to do a direct-io. Reading the code i figured it used 'get_user_pages' and since VRAM is io mapped, those pages are marked VM_IO and hence directio fai