[PATCH v4] filemap: add trace events for get_pages, map_pages, and fault

2024-08-16 Thread Takaya Saeki
he performance of boot up. With the combination of these 3 new trace events and mm_filemap_add_to_page_cache, we observed a reduction in the pack file by 7.3% - 20% on ChromeOS varying by device. Signed-off-by: Takaya Saeki Reviewed-by: Masami Hiramatsu (Google) Reviewed-by: Steven Rostedt (Google)

[PATCH v3] filemap: add trace events for get_pages, map_pages, and fault

2024-08-13 Thread Takaya Saeki
he performance of boot up. With the combination of these 3 new trace events and mm_filemap_add_to_page_cache, we observed a reduction in the pack file by 7.3% - 20% on ChromeOS varying by device. Signed-off-by: Takaya Saeki Reviewed-by: Masami Hiramatsu (Google) Reviewed-by: Steven Rostedt (Google)

Re: [PATCH v2] filemap: add trace events for get_pages, map_pages, and fault

2024-08-08 Thread Takaya Saeki
Oh yes. That sounds like a good idea. Let me update the patch with it. Thanks, Takaya Saeki

Re: [PATCH v2] filemap: add trace events for get_pages, map_pages, and fault

2024-07-10 Thread Takaya Saeki
Hello Matthew, I'd appreciate it if you could comment on this. Thank you.

Re: [PATCH v2] filemap: add trace events for get_pages, map_pages, and fault

2024-07-02 Thread Takaya Saeki
Hello all, and thank you so much for the review, Steven and Masami. I'm currently considering replacing the `max_ofs` output with `length`. Please let me know your thoughts. With the current design, a memory range of an event is an inclusive range of [ofs, max_ofs + 4096]. I found the `+4096` part

[PATCH v2] filemap: add trace events for get_pages, map_pages, and fault

2024-06-20 Thread Takaya Saeki
he performance of boot up. With the combination of these 3 new trace events and mm_filemap_add_to_page_cache, we observed a reduction in the pack file by 7.3% - 20% on ChromeOS varying by device. Signed-off-by: Takaya Saeki --- Changelog between v2 and v1 - Fix a file offset type usage by casting

Re: [PATCH] filemap: add trace events for get_pages, map_pages, and fault

2024-06-20 Thread Takaya Saeki
Thank you Matthew for taking a look at this! I fix both points and send a V2 patch. > This needs to be cast to an loff_t before shifting. I found that this applies to add_to_page_cache and delete_from_page_cache, too. I could fix them too, but I didn't since it will change the max value of those

Re: [PATCH] filemap: add trace events for get_pages, map_pages, and fault

2024-06-19 Thread Takaya Saeki
Thanks Steven for trying it out. > I can see it bringing down the number of pages needed to be saved > dramatically. Yes, I agree. However, note that wc does not count the size of the page caches correctly since 'get_map_pages' gives you a range. In your example of the less command, actually the

[PATCH] filemap: add trace events for get_pages, map_pages, and fault

2024-06-18 Thread Takaya Saeki
he performance of boot up. With the combination of these 3 new trace events and mm_filemap_add_to_page_cache, we observed a reduction in the pack file by 7.3% - 20% on ChromeOS varying by device. Signed-off-by: Takaya Saeki --- include/trace/events/filemap.h | 84 +++