From: Sascha Hauer <s.hauer@pengutronix.de>
To: Ahmad Fatoum <a.fatoum@pengutronix.de>
Cc: barebox@lists.infradead.org
Subject: Re: [PATCH v2 08/10] mci: add support for discarding write blocks
Date: Thu, 1 Aug 2024 13:27:51 +0200 [thread overview]
Message-ID: <ZqtxNwCHEdv5VIFP@pengutronix.de> (raw)
In-Reply-To: <20240731080510.364706-9-a.fatoum@pengutronix.de>
On Wed, Jul 31, 2024 at 10:05:08AM +0200, Ahmad Fatoum wrote:
> +/**
> + * Erase a memory region
> + * @param blk All info about the block device we need
> + * @param block first block to erase
> + * @param num_blocks Number of blocks to erase
> + * @return 0 on success, anything else on failure
> + *
> + */
> +static int mci_sd_erase(struct block_device *blk, sector_t from,
> + blkcnt_t blkcnt)
> +{
> + struct mci_part *part = container_of(blk, struct mci_part, blk);
> + struct mci *mci = part->mci;
> + sector_t i = 0;
> + unsigned arg;
> + sector_t to = from + blkcnt;
> + int rc;
> +
> + mci_blk_part_switch(part);
> +
> + rc = mci_sd_check_write(mci, "Erase", from, blkcnt);
> + if (rc)
> + return rc;
> +
> + if (!mci->erase_grp_size)
> + return -EOPNOTSUPP;
> +
> + if (mci->can_trim) {
> + arg = MMC_TRIM_ARG;
> + } else {
> + /* We don't use discard, as it doesn't guarantee a fixed value */
> + arg = MMC_ERASE_ARG;
> + blkcnt = mmc_align_erase_size(mci, &from, &to, blkcnt);
> + }
> +
> + if (blkcnt == 0)
> + return 0;
> +
> + /* 'from' and 'to' are inclusive */
> + to -= 1;
> +
> + while (i < blkcnt) {
> + sector_t blk_r;
> +
> + /* TODO: While it's possible to clear many erase groups at once
> + * and it greatly improves throughput, drivers need adjustment:
> + *
> + * Many drivers hardcode a maximal wait time before aborting
> + * the wait for R1b and returning -ETIMEDOUT. With long
> + * erases/trims, we are bound to run into this timeout, so for now
> + * we just split into sufficiently small erases that are unlikely
> + * to trigger the timeout.
> + *
> + * What Linux does and what we should be doing in barebox is:
> + *
> + * - add a struct mci_cmd::busy_timeout member that drivers should
> + * use instead of hardcoding their own timeout delay. The busy
> + * timeout length can be calculated by the MCI core after
> + * consulting the appropriate CSD/EXT_CSD/SSR registers.
> + *
> + * - add a struct mci_host::max_busy_timeout member, where drivers
> + * can indicate the maximum timeout they are able to support.
> + * The MCI core will never set a busy_timeout that exceeds this
> + * value.
> + *
> + * Example Samsung eMMC 8GTF4:
> + *
> + * time erase /dev/mmc2.part_of_512m # 1024 trims
> + * time: 2849ms
> + *
> + * time erase /dev/mmc2.part_of_512m # single trim
> + * time: 56ms
> + */
> +
> + if (IS_SD(mci) && mci->ssr.au) {
> + blk_r = ((blkcnt - i) > mci->ssr.au) ?
> + mci->ssr.au : (blkcnt - i);
> + } else {
> + blk_r = ((blkcnt - i) > mci->erase_grp_size) ?
> + mci->erase_grp_size : (blkcnt - i);
> + }
The number of blocks we can erase in one iteration is constant, I pulled
this out of the loop. Please check the result in -next.
Sascha
--
Pengutronix e.K. | |
Steuerwalder Str. 21 | http://www.pengutronix.de/ |
31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
next prev parent reply other threads:[~2024-08-01 11:30 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-07-31 8:05 [PATCH v2 00/10] mmc: add SD/eMMC erase support Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 01/10] fs: give erase() a new erase_type parameter Ahmad Fatoum
2024-08-01 10:51 ` Sascha Hauer
2024-08-01 10:54 ` Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 02/10] cdev: factor out range identical/overlap check Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 03/10] block: factor out chunk_flush helper Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 04/10] block: allow block devices to implement the cdev erase operation Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 05/10] mci: turn bool members into bitfield in struct mci Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 06/10] mci: describe more command structure in mci.h Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 07/10] mci: core: use CONFIG_MCI_WRITE, not CONFIG_BLOCK_WRITE Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 08/10] mci: add support for discarding write blocks Ahmad Fatoum
2024-08-01 11:27 ` Sascha Hauer [this message]
2024-07-31 8:05 ` [PATCH v2 09/10] commands: sync: add new command to flush cached writes Ahmad Fatoum
2024-07-31 8:05 ` [PATCH v2 10/10] commands: blkstats: add command to print block device statistics Ahmad Fatoum
2024-08-01 10:49 ` [PATCH v2 00/10] mmc: add SD/eMMC erase support Sascha Hauer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZqtxNwCHEdv5VIFP@pengutronix.de \
--to=s.hauer@pengutronix.de \
--cc=a.fatoum@pengutronix.de \
--cc=barebox@lists.infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox