[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAGWkznHq15QN5Dn6_QfbAm7jS9OPCV4TVqn2_9RxUBx0V9v78w@mail.gmail.com>
Date: Wed, 1 Dec 2021 19:12:30 +0800
From: Zhaoyang Huang <huangzhaoyang@...il.com>
To: Nitin Gupta <ngupta@...are.org>,
Sergey Senozhatsky <senozhatsky@...omium.org>,
Jens Axboe <axboe@...nel.dk>,
Johannes Weiner <hannes@...xchg.org>,
Minchan Kim <minchan@...nel.org>,
Zhaoyang Huang <zhaoyang.huang@...soc.com>,
"open list:MEMORY MANAGEMENT" <linux-mm@...ck.org>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [RFC PATCH] mm: count zram read/write into PSI_IO_WAIT
There is no chance for zram reading/writing to be counted in
PSI_IO_WAIT so far as zram will deal with the request just in current
context without invoking submit_bio and io_schedule.
On Wed, Dec 1, 2021 at 6:59 PM Huangzhaoyang <huangzhaoyang@...il.com> wrote:
>
> From: Zhaoyang Huang <zhaoyang.huang@...soc.com>
>
> Have zram reading/writing be counted in PSI_IO_WAIT.
>
> Signed-off-by: Zhaoyang Huang <zhaoyang.huang@...soc.com>
> ---
> drivers/block/zram/zram_drv.c | 18 ++++++++++++++++--
> 1 file changed, 16 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
> index fcaf275..b0e4766 100644
> --- a/drivers/block/zram/zram_drv.c
> +++ b/drivers/block/zram/zram_drv.c
> @@ -34,6 +34,7 @@
> #include <linux/debugfs.h>
> #include <linux/cpuhotplug.h>
> #include <linux/part_stat.h>
> +#include <linux/psi.h>
>
> #include "zram_drv.h"
>
> @@ -1246,7 +1247,9 @@ static int __zram_bvec_read(struct zram *zram, struct page *page, u32 index,
> zram_get_element(zram, index),
> bio, partial_io);
> }
> -
> +#ifdef CONFIG_PSI
> + psi_task_change(current, 0, TSK_IOWAIT);
> +#endif
> handle = zram_get_handle(zram, index);
> if (!handle || zram_test_flag(zram, index, ZRAM_SAME)) {
> unsigned long value;
> @@ -1257,6 +1260,9 @@ static int __zram_bvec_read(struct zram *zram, struct page *page, u32 index,
> zram_fill_page(mem, PAGE_SIZE, value);
> kunmap_atomic(mem);
> zram_slot_unlock(zram, index);
> +#ifdef CONFIG_PSI
> + psi_task_change(current, TSK_IOWAIT, 0);
> +#endif
> return 0;
> }
>
> @@ -1284,6 +1290,9 @@ static int __zram_bvec_read(struct zram *zram, struct page *page, u32 index,
> if (WARN_ON(ret))
> pr_err("Decompression failed! err=%d, page=%u\n", ret, index);
>
> +#ifdef CONFIG_PSI
> + psi_task_change(current, TSK_IOWAIT, 0);
> +#endif
> return ret;
> }
>
> @@ -1471,7 +1480,13 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec,
> vec.bv_offset = 0;
> }
>
> +#ifdef CONFIG_PSI
> + psi_task_change(current, 0, TSK_IOWAIT);
> +#endif
> ret = __zram_bvec_write(zram, &vec, index, bio);
> +#ifdef CONFIG_PSI
> + psi_task_change(current, TSK_IOWAIT, 0);
> +#endif
> out:
> if (is_partial_io(bvec))
> __free_page(page);
> @@ -1607,7 +1622,6 @@ static blk_qc_t zram_submit_bio(struct bio *bio)
> atomic64_inc(&zram->stats.invalid_io);
> goto error;
> }
> -
> __zram_make_request(zram, bio);
> return BLK_QC_T_NONE;
>
> --
> 1.9.1
>
Powered by blists - more mailing lists