Thread (2 messages) flat view 2 messages, 2 authors, 4d ago

Re: [PATCH] fbdev: omapfb: Fix __be32 sparse warning in panel_enabled()

From: Helge Deller <deller@gmx.de>
Date: 2026-08-22 10:23:45
Also in: dri-devel, linux-omap, sashiko-reviews

CC'ing the linux-omap mailing list, in the hope someone with omap
knowledge can judge if the sashiko AI bot is right with his analysis that
acx565akm_transfer() might be wrong on big-endian systems?

On 8/22/26 11:58, sashiko-bot@kernel.org wrote:
quoted hunk
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Hardcoded little-endian assumption when transmitting the command byte breaks the driver on big-endian architectures.
--
--- Patch [1]: [PATCH] fbdev: omapfb: Fix __be32 sparse warning in panel_enabled() ---
commit a9e06be56fc1b74d7c6e76133e9ea1c25675a615
Author: Helge Deller [off-list ref]
Subject: fbdev: omapfb: Fix __be32 sparse warning in panel_enabled()

This commit resolves a sparse warning in panel_enabled() by explicitly
reading the display status into a __be32 variable. It then converts this
value to CPU endianness using __be32_to_cpu() before checking the bits.

Closes: https://lore.kernel.org/oe-kbuild-all/202608211811.lps93xao-lkp@intel.com/ (local)
quoted
diff --git a/drivers/video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c b/drivers/video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c
index 8f430d9e80549..0202ca8cbfc23 100644
--- a/drivers/video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c
+++ b/drivers/video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c
@@ -210,12 +210,13 @@ static void set_display_state(struct panel_drv_data *ddata, int enabled)
  
  static int panel_enabled(struct panel_drv_data *ddata)
  {
+	__be32 disp_status_be;
  	u32 disp_status;
  	int enabled;
  
  	acx565akm_read(ddata, MIPID_CMD_READ_DISP_STATUS,
-			(u8 *)&disp_status, 4);
[Severity: High]
This is a pre-existing issue, but while looking at endianness fixes in this
file, does acx565akm_transfer() have a similar issue on big-endian
architectures?

When acx565akm_read() calls acx565akm_transfer(), it passes the command as
an int:

drivers/video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c:acx565akm_transfer() {
     ...
     cmd &=  0xff;
     x->tx_buf = &cmd;
     x->bits_per_word = 9;
     x->len = 2;
     ...
}

Because bits_per_word > 8, the SPI controller reads the buffer as a 16-bit
integer. Taking the address of a 32-bit int on a little-endian system yields
the command byte in the lowest bits. However, on big-endian systems, &cmd
points to the most significant bytes (which are 0x00).

Will the SPI controller end up transmitting 0x0000 instead of the requested
command on big-endian architectures?
quoted
-	disp_status = __be32_to_cpu(disp_status);
+			(u8 *)&disp_status_be, 4);
+	disp_status = __be32_to_cpu(disp_status_be);
  	enabled = (disp_status & (1 << 17)) && (disp_status & (1 << 10));
  	dev_dbg(&ddata->spi->dev,
  		"LCD panel %senabled by bootloader (status 0x%04x)\n",
  		enabled ? "" : "not ", disp_status);
  
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help