mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Ahmad Fatoum <a.fatoum@pengutronix.de>
To: barebox@lists.infradead.org
Cc: Ahmad Fatoum <a.fatoum@pengutronix.de>
Subject: [PATCH 08/10] commands: sync: add new command to flush cached writes
Date: Tue, 30 Jul 2024 09:19:27 +0200	[thread overview]
Message-ID: <20240730071929.2016537-9-a.fatoum@pengutronix.de> (raw)
In-Reply-To: <20240730071929.2016537-1-a.fatoum@pengutronix.de>

In barebox, the dirty block cache for a given device is flushed when
closing all file descriptors to it, a cached block needs to be evicted
due to pressure or when barebox shuts down.

There are valid uses to do it explicitly however:

  - When benchmarking write speed to a block device
  - Before forced resets, because a regular reset would run shutdown
    code that should be omitted.

Add a sync command that does this. Unlike the userspace variant, we
don't both to implement a argument-less sync that's global. Users are
supposed to flush the files they are actually interested in, e.g.:

  sync /dev/mmc2.0

Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
---
 commands/Kconfig  | 14 ++++++++++++++
 commands/Makefile |  1 +
 commands/sync.c   | 42 ++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 57 insertions(+)
 create mode 100644 commands/sync.c

diff --git a/commands/Kconfig b/commands/Kconfig
index 5b512f1bbac7..a8b7037618cc 100644
--- a/commands/Kconfig
+++ b/commands/Kconfig
@@ -1871,6 +1871,20 @@ config CMD_DETECT
 		  -e	bail out if one device fails to detect
 		  -a	detect all devices
 
+config CMD_SYNC
+	tristate
+	prompt "sync"
+	help
+	  sync - flush cached writes
+
+	  Usage: sync DEVICE
+
+	  Synchronize cached writes to persistent storage of DEVICE
+	  immediately instead of waiting for block device to be closed,
+	  cache eviction or the regular barebox exit.
+
+	  This command is mainly useful for benchmarking.
+
 config CMD_FLASH
 	tristate
 	prompt "erase, protect and unprotect"
diff --git a/commands/Makefile b/commands/Makefile
index 4ca7ba7eb609..a9dbead4389f 100644
--- a/commands/Makefile
+++ b/commands/Makefile
@@ -41,6 +41,7 @@ obj-$(CONFIG_CMD_REGINFO)	+= reginfo.o
 obj-$(CONFIG_CMD_CRC)		+= crc.o
 obj-$(CONFIG_CMD_CLEAR)		+= clear.o
 obj-$(CONFIG_CMD_TEST)		+= test.o
+obj-$(CONFIG_CMD_SYNC)		+= sync.o
 obj-$(CONFIG_CMD_FLASH)		+= flash.o
 obj-$(CONFIG_CMD_MEMINFO)	+= meminfo.o
 obj-$(CONFIG_CMD_TIMEOUT)	+= timeout.o
diff --git a/commands/sync.c b/commands/sync.c
new file mode 100644
index 000000000000..6add67cd0876
--- /dev/null
+++ b/commands/sync.c
@@ -0,0 +1,42 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+/* sync - flushing support */
+
+#include <common.h>
+#include <command.h>
+#include <fcntl.h>
+#include <fs.h>
+
+static int do_sync(int argc, char *argv[])
+{
+	int ret, fd;
+
+	if (argc != 2)
+		return COMMAND_ERROR_USAGE;
+
+	fd = open(argv[1], O_WRONLY);
+	if (fd < 0) {
+		printf("open %s: %m\n", argv[1]);
+		return 1;
+	}
+
+	ret = flush(fd);
+
+	close(fd);
+
+	return ret;
+}
+
+BAREBOX_CMD_HELP_START(sync)
+BAREBOX_CMD_HELP_TEXT("Synchronize cached writes to persistent storage of DEVICE")
+BAREBOX_CMD_HELP_TEXT("immediately instead of waiting for block device to be closed,")
+BAREBOX_CMD_HELP_TEXT("cache eviction or the regular barebox exit.")
+BAREBOX_CMD_HELP_END
+
+BAREBOX_CMD_START(sync)
+	.cmd		= do_sync,
+	BAREBOX_CMD_DESC("flush cached writes")
+	BAREBOX_CMD_OPTS("DEVICE")
+	BAREBOX_CMD_GROUP(CMD_GRP_HWMANIP)
+	BAREBOX_CMD_HELP(cmd_sync_help)
+BAREBOX_CMD_END
-- 
2.39.2




  parent reply	other threads:[~2024-07-30  7:20 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-07-30  7:19 [PATCH 00/10] mmc: add SD/eMMC erase support Ahmad Fatoum
2024-07-30  7:19 ` [PATCH 01/10] fs: give erase() a new erase_type parameter Ahmad Fatoum
2024-07-30  8:31   ` Sascha Hauer
2024-07-30  8:32     ` Ahmad Fatoum
2024-07-30  7:19 ` [PATCH 02/10] block: factor out chunk_flush helper Ahmad Fatoum
2024-07-30  7:19 ` [PATCH 03/10] block: allow block devices to implement the cdev erase operation Ahmad Fatoum
2024-07-30  8:55   ` Sascha Hauer
2024-07-30 11:10     ` Ahmad Fatoum
2024-07-30 11:21       ` Sascha Hauer
2024-07-30  7:19 ` [PATCH 04/10] mci: turn bool members into bitfield in struct mci Ahmad Fatoum
2024-07-30  9:06   ` Sascha Hauer
2024-07-30 11:10     ` Ahmad Fatoum
2024-07-30  7:19 ` [PATCH 05/10] mci: describe more command structure in mci.h Ahmad Fatoum
2024-07-30  9:25   ` Yann Sionneau
2024-07-30 11:07     ` Ahmad Fatoum
2024-07-30  7:19 ` [PATCH 06/10] mci: core: use CONFIG_MCI_WRITE, not CONFIG_BLOCK_WRITE Ahmad Fatoum
2024-07-30  9:18   ` Sascha Hauer
2024-07-30 11:08     ` Ahmad Fatoum
2024-07-31  7:19       ` Ahmad Fatoum
2024-07-30  7:19 ` [PATCH 07/10] mci: add support for discarding write blocks Ahmad Fatoum
2024-07-30  9:23   ` Yann Sionneau
2024-07-30 11:14     ` Ahmad Fatoum
2024-07-30 10:05   ` Sascha Hauer
2024-07-30 11:17     ` Ahmad Fatoum
2024-07-30  7:19 ` Ahmad Fatoum [this message]
2024-07-30 10:08   ` [PATCH 08/10] commands: sync: add new command to flush cached writes Sascha Hauer
2024-07-30  7:19 ` [PATCH 09/10] mci: core: remove reference to SD Card from common mci_card_probe Ahmad Fatoum
2024-07-30 10:09   ` Sascha Hauer
2024-07-30  7:19 ` [PATCH 10/10] commands: blkstats: add command to print block device statistics Ahmad Fatoum

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=20240730071929.2016537-9-a.fatoum@pengutronix.de \
    --to=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