[PATCH] ata: libata-core: Ignore forced transfer modes the host does not support

[PATCH] ata: libata-core: Ignore forced transfer modes the host does not support

[Rand Deeb]

The libata.force parameter can force a transfer mode (pio0-6, mwdma0-4,
udma0-7) for a port or device. ata_force_xfermask() applies it in
ata_do_set_mode() right after ata_dev_xfermask() has limited the device
masks to what the host supports. It replaces those masks with the forced
mode without checking the host masks, so a mode the controller does not
support is passed to the driver's ->set_piomode() or ->set_dmamode().

Several PATA drivers rely on the host masks to keep the mode number in
range when indexing their timing tables. ata_piix, pata_efar,
pata_it8213, pata_oldpiix, pata_radisys and pata_rdc declare ATA_PIO4
and ATA_MWDMA12_ONLY, index a five-entry PIO timing table with
pio_mode - XFER_PIO_0, and index a three-entry needed_pio[] table with
dma_mode - XFER_MW_DMA_0. Booting with libata.force=pio5, pio6, mwdma3
or mwdma4 makes them read past those tables, use the values read as
further indices into the timing table, and write the result to the
controller's timing registers.

Make ata_force_xfermask() ignore a forced transfer mode that is not in
the port's PIO, MWDMA or UDMA mask, and warn about it. Forcing a mode the
host supports still works as before, including a mode that
ata_dev_xfermask() removed for the device or cable.

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Fixes: 332673257056 ("libata: implement libata.force module parameter")
Signed-off-by: Rand Deeb <rand.sec96@gmail.com>
---
 drivers/ata/libata-core.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 92233fb25..324aaa81c 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -434,6 +434,7 @@ static void ata_force_link_limits(struct ata_link *link)
  */
 static void ata_force_xfermask(struct ata_device *dev)
 {
+	struct ata_port *ap = dev->link->ap;
 	int devno = dev->link->pmp + dev->devno;
 	int alt_devno = devno;
 	int i;
@@ -458,6 +459,17 @@ static void ata_force_xfermask(struct ata_device *dev)
 
 		ata_unpack_xfermask(fe->param.xfer_mask,
 				    &pio_mask, &mwdma_mask, &udma_mask);
+
+		/* Never force a transfer mode the host does not support */
+		if ((pio_mask & ~ap->pio_mask) ||
+		    (mwdma_mask & ~ap->mwdma_mask) ||
+		    (udma_mask & ~ap->udma_mask)) {
+			ata_dev_warn(dev,
+				     "FORCE: %s not supported by the host, ignored\n",
+				     fe->param.name);
+			return;
+		}
+
 		if (udma_mask)
 			dev->udma_mask = udma_mask;
 		else if (mwdma_mask) {

base-commit: 4056b3c9d3c92ea45955641dbafeb9b960e146ef
-- 
2.39.5

Re: [PATCH] ata: libata-core: Ignore forced transfer modes the host does not support

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] ata: libata-core: early return bypasses valid fallback force configs
--

commit b2b8699db9f1c2c52ee37c5d556e07a3dde1ab2e
Author: Rand Deeb <rand.sec96@gmail.com>

ata: libata-core: Ignore forced transfer modes the host does not support

This commit adds a check to ignore forced transfer modes that the host
controller does not support. This prevents out-of-bounds reads in driver
timing tables when an unsupported mode is forced.

> diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
> index 92233fb25051a..324aaa81c8a30 100644
> --- a/drivers/ata/libata-core.c
> +++ b/drivers/ata/libata-core.c
> @@ -458,6 +459,17 @@ static void ata_force_xfermask(struct ata_device *dev)
>  
>  		ata_unpack_xfermask(fe->param.xfer_mask,
>  				    &pio_mask, &mwdma_mask, &udma_mask);
> +
> +		/* Never force a transfer mode the host does not support */
> +		if ((pio_mask & ~ap->pio_mask) ||
> +		    (mwdma_mask & ~ap->mwdma_mask) ||
> +		    (udma_mask & ~ap->udma_mask)) {
> +			ata_dev_warn(dev,
> +				     "FORCE: %s not supported by the host, ignored\n",
> +				     fe->param.name);
> +			return;

[Severity: Medium]
When a specific override is rejected by this new host capability check in
ata_force_xfermask(), does using return prematurely abort the loop?

The libata.force parameter allows stacking multiple hierarchical overrides
(e.g., libata.force=pio4,1.0:pio5), which are evaluated backwards so the
most specific match wins.

By returning instead of using continue, the code appears to silently bypass
earlier valid fallback overrides like the port-wide pio4 in the example above.

This would cause the device to fall back to its auto-negotiated mask instead
of the valid user-specified fallback, which might cause issues if the user
was relying on the fallback constraint to disable broken DMA.

> +		}
> +
>  		if (udma_mask)
>  			dev->udma_mask = udma_mask;
>  		else if (mwdma_mask) {

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261014125916.1975-1-rand.sec96@gmail.com?part=1