From: "Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>
To: chitranshmandhaniya@gmail.com
Cc: Hans de Goede <hansg@kernel.org>,
platform-driver-x86@vger.kernel.org,
LKML <linux-kernel@vger.kernel.org>,
Chitransh Mandhaniya <Chitransh_Mandhaniya@dell.com>
Subject: Re: [PATCH] platform/x86: dell-uart-backlight: recover from stray EC response byte
Date: Wed, 14 Oct 2026 14:50:01 +0300 (EEST) [thread overview]
Message-ID: <84756da9-dab7-84df-0469-dadfedb1eb78@linux.intel.com> (raw)
In-Reply-To: <20261004110333.5790-1-chitranshmandhaniya@gmail.com>
On Sun, 4 Oct 2026, chitranshmandhaniya@gmail.com wrote:
> From: Chitransh Mandhaniya <Chitransh_Mandhaniya@dell.com>
>
> On some Dell AIO machines the EC occasionally inserts a spurious byte
> after the cmd-echo in UART backlight response frames. A get-brightness
> reply "04 0c 0a e5" arrives as "04 0c 0d 0a e5". The length field still
> reflects the original frame size, so the parser consumes the stray byte
> as payload and mistakes the real payload for the checksum, returning
> -EIO.
>
> Handle this in dell_uart_bl_receive(): on the first checksum failure for
> a frame longer than the minimum size, drop the byte at the first payload
> offset, shift the rest down by one, read one more byte as the new
> checksum, and re-validate. Limit this single-byte resync to once per
> command via a resync_used flag cleared at the start of each transaction.
> A second mismatch is still a hard error.
>
> Well-formed replies pass on the first check and never enter the recovery
> path.
>
> Signed-off-by: Chitransh Mandhaniya <Chitransh_Mandhaniya@dell.com>
> ---
> .../platform/x86/dell/dell-uart-backlight.c | 92 +++++++++++++------
> 1 file changed, 62 insertions(+), 30 deletions(-)
>
> diff --git a/drivers/platform/x86/dell/dell-uart-backlight.c b/drivers/platform/x86/dell/dell-uart-backlight.c
> index f323a667dc2d..068b91c21bf4 100644
> --- a/drivers/platform/x86/dell/dell-uart-backlight.c
> +++ b/drivers/platform/x86/dell/dell-uart-backlight.c
> @@ -51,7 +51,7 @@
> #define MAX_RESP_LEN 80
>
> struct dell_uart_backlight {
> - struct mutex mutex;
> + struct mutex mutex; /* Protects command + response state below */
> wait_queue_head_t wait_queue;
> struct device *dev;
> struct backlight_device *bl;
> @@ -60,6 +60,7 @@ struct dell_uart_backlight {
> u8 resp_len;
> u8 resp_max_len;
> u8 pending_cmd;
> + bool resync_used;
> int status;
> int power;
> };
> @@ -91,6 +92,7 @@ static int dell_uart_bl_command(struct dell_uart_backlight *dell_bl,
> dell_bl->resp_len = -1; /* Invalid / unset */
> dell_bl->resp_max_len = resp_max_len;
> dell_bl->pending_cmd = cmd[1];
> + dell_bl->resync_used = false;
>
> /* The TTY buffer should be big enough to take the entire cmd in one go */
> ret = serdev_device_write_buf(to_serdev_device(dell_bl->dev), cmd, cmd_len);
> @@ -208,8 +210,10 @@ static const struct backlight_ops dell_uart_backlight_ops = {
> static size_t dell_uart_bl_receive(struct serdev_device *serdev, const u8 *data, size_t len)
> {
> struct dell_uart_backlight *dell_bl = serdev_device_get_drvdata(serdev);
> + const int max_passes = 2;
> size_t i;
> u8 csum;
> + int pass;
>
> dev_dbg(dell_bl->dev, "Recv: %*ph\n", (int)len, data);
>
> @@ -220,42 +224,70 @@ static size_t dell_uart_bl_receive(struct serdev_device *serdev, const u8 *data,
> }
>
> i = 0;
> - while (i < len && dell_bl->resp_idx != dell_bl->resp_len) {
> - dell_bl->resp[dell_bl->resp_idx] = data[i++];
> -
> - switch (dell_bl->resp_idx) {
> - case RESP_LEN: /* Length byte */
> - dell_bl->resp_len = dell_bl->resp[RESP_LEN];
> - if (dell_bl->resp_len < MIN_RESP_LEN ||
> - dell_bl->resp_len > dell_bl->resp_max_len) {
> - dev_err(dell_bl->dev, "Response length %d out if range %d - %d\n",
> - dell_bl->resp_len, MIN_RESP_LEN, dell_bl->resp_max_len);
> - dell_bl->status = -EIO;
> - goto wakeup;
> + for (pass = 0; pass < max_passes; pass++) {
> + while (i < len && dell_bl->resp_idx != dell_bl->resp_len) {
Hi,
Could the inner part be moved into a function so you wouldn't need to add
the pass based approach but could just call the function again? I think
that would be much easier to follow than adding a loop for retry.
If you end up doing that, please make the function refactoring in a
separate change before adding the extra byte handling.
--
i.
> + dell_bl->resp[dell_bl->resp_idx] = data[i++];
> +
> + switch (dell_bl->resp_idx) {
> + case RESP_LEN: /* Length byte */
> + dell_bl->resp_len = dell_bl->resp[RESP_LEN];
> + if (dell_bl->resp_len < MIN_RESP_LEN ||
> + dell_bl->resp_len > dell_bl->resp_max_len) {
> + dev_err(dell_bl->dev,
> + "Response length %d out if range %d - %d\n",
> + dell_bl->resp_len, MIN_RESP_LEN,
> + dell_bl->resp_max_len);
> + dell_bl->status = -EIO;
> + goto wakeup;
> + }
> + break;
> + case RESP_CMD: /* CMD byte */
> + if (dell_bl->resp[RESP_CMD] != dell_bl->pending_cmd) {
> + dev_err(dell_bl->dev, "Response cmd 0x%02x != pending 0x%02x\n",
> + dell_bl->resp[RESP_CMD], dell_bl->pending_cmd);
> + dell_bl->status = -EIO;
> + goto wakeup;
> + }
> + break;
> }
> - break;
> - case RESP_CMD: /* CMD byte */
> - if (dell_bl->resp[RESP_CMD] != dell_bl->pending_cmd) {
> - dev_err(dell_bl->dev, "Response cmd 0x%02x != pending 0x%02x\n",
> - dell_bl->resp[RESP_CMD], dell_bl->pending_cmd);
> - dell_bl->status = -EIO;
> - goto wakeup;
> - }
> - break;
> + dell_bl->resp_idx++;
> }
> - dell_bl->resp_idx++;
> - }
>
> - if (dell_bl->resp_idx != dell_bl->resp_len)
> - return len; /* Response not complete yet */
> + if (dell_bl->resp_idx != dell_bl->resp_len)
> + return i; /* Response not complete yet, wait for more data */
> +
> + csum = dell_uart_checksum(dell_bl->resp, dell_bl->resp_len - 1);
> + if (dell_bl->resp[dell_bl->resp_len - 1] == csum) {
> + dell_bl->status = 0; /* Success */
> + dev_dbg(dell_bl->dev,
> + "Response OK for cmd 0x%02x, len %d: %*ph\n",
> + dell_bl->pending_cmd, dell_bl->resp_len,
> + dell_bl->resp_len, dell_bl->resp);
> + goto wakeup;
> + }
> +
> + /*
> + * Some EC firmware inserts a stray byte right after the cmd-echo
> + * byte. Try once to drop it and re-fetch the checksum byte from the wire.
> + */
> + if (!dell_bl->resync_used && dell_bl->resp_len > MIN_RESP_LEN) {
> + dell_bl->resync_used = true;
> +
> + dev_warn(dell_bl->dev,
> + "Checksum mismatch got 0x%02x expected 0x%02x, dropping stray byte 0x%02x at offset %d\n",
> + dell_bl->resp[dell_bl->resp_len - 1], csum,
> + dell_bl->resp[RESP_DATA], RESP_DATA);
> +
> + memmove(&dell_bl->resp[RESP_DATA], &dell_bl->resp[RESP_DATA + 1],
> + dell_bl->resp_len - RESP_DATA - 1);
> + dell_bl->resp_idx = dell_bl->resp_len - 1;
> + continue; /* re-fetch the last (checksum) byte and re-validate */
> + }
>
> - csum = dell_uart_checksum(dell_bl->resp, dell_bl->resp_len - 1);
> - if (dell_bl->resp[dell_bl->resp_len - 1] == csum) {
> - dell_bl->status = 0; /* Success */
> - } else {
> dev_err(dell_bl->dev, "Checksum mismatch got 0x%02x expected 0x%02x\n",
> dell_bl->resp[dell_bl->resp_len - 1], csum);
> dell_bl->status = -EIO;
> + goto wakeup;
> }
> wakeup:
> wake_up(&dell_bl->wait_queue);
>
--
i.
next prev parent reply other threads:[~2026-10-14 11:50 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-04 11:03 chitranshmandhaniya
2026-10-14 11:50 ` Ilpo Järvinen [this message]
2026-10-15 7:43 ` [PATCH v2 0/2] " chitranshmandhaniya
2026-10-15 7:43 ` [PATCH v2 1/2] platform/x86: dell-uart-backlight: extract dell_uart_bl_receive_data() helper chitranshmandhaniya
2026-10-15 7:43 ` [PATCH v2 2/2] platform/x86: dell-uart-backlight: recover from stray EC response byte chitranshmandhaniya
2026-10-15 10:12 ` Ilpo Järvinen
2026-10-15 10:24 ` Ilpo Järvinen
2026-10-16 9:33 ` [PATCH v3 0/2] platform/x86: dell-uart-backlight: stray byte recovery and cleanup chitranshmandhaniya
2026-10-16 9:33 ` [PATCH v3 1/2] platform/x86: dell-uart-backlight: extract chitranshmandhaniya
2026-10-16 9:33 ` [PATCH v3 2/2] platform/x86: dell-uart-backlight: recover from stray chitranshmandhaniya
2026-10-16 10:07 ` Ilpo Järvinen
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=84756da9-dab7-84df-0469-dadfedb1eb78@linux.intel.com \
--to=ilpo.jarvinen@linux.intel.com \
--cc=Chitransh_Mandhaniya@dell.com \
--cc=chitranshmandhaniya@gmail.com \
--cc=hansg@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=platform-driver-x86@vger.kernel.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
all inboxes | Powered by JetHome®