Skip to contentSkip to conversation

clk: rs9: Fix and of_clk_hw_onecell_get() conversion

PATCH Applied Geert Uytterhoeven 3 participants 5 messages Sep 3 – Sep 11 Common clk

lore
Share & tools
lore
#0 lore raw

v5 cover clk: rs9: Fix and of_clk_hw_onecell_get() conversion

Hi all,

This patch series contains a fix and further development for the Renesas 9-series PCIe clock generator driver, which were submitted before as separate patches As the fix hasn't been picked up yet, and the further development depends on the fix, I am resending them as a series.

Changes compared to v4[1]:
- Add Reviewed-by.

Thanks for your comments!

[1] "[PATCH v4 0/2] clk: rs9: Fix and of_clk_hw_onecell_get() conversion"

    https://lore.kernel.org/r/[email protected]

Geert Uytterhoeven (2):
clk: rs9: Add clock index range check to rs9_of_clk_get()
clk: rs9: Convert to clk_hw_onecell_data and of_clk_hw_onecell_get()

 drivers/clk/clk-renesas-pcie.c | 39 ++++++++++++++++------------------
 1 file changed, 18 insertions(+), 21 deletions(-)
Signature · 12 lines
2.43.0

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds
#1 lore raw

PATCH v5 1/2 clk: rs9: Add clock index range check to rs9_of_clk_get()

rs9_of_clk_get() does not validate the clock index in the passed DT clock specifier. If DT specifies an incorrect and out-of-range index, this will access memory beyond the end of the clk_dif[] array.

Fix by this adding a range check to rs9_of_clk_get().

Fixes 892e0ddea1aa6f70 ("clk: rs9: Add Renesas 9-series PCIe clock generator driver")

---
v5:
- Add Reviewed-by,

v4:
- Put in a series,

v3:
- Add error message,

v2:
- Just add the missing range check; the conversion to of_clk_hw_onecell_get() can be done later.

v1: "[PATCH] clk: rs9: Convert to clk_hw_onecell_data and

     of_clk_hw_onecell_get()"
    https://lore.kernel.org/a6dce17b15d29a257d09fe0edc199a14c297f1a8.1768836042.git.geert+renesas@glider.be
---
 drivers/clk/clk-renesas-pcie.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/clk/clk-renesas-pcie.c +5 -0
index 564550c803896756..f3376c29007aa92d 100644
@@ -276,6 +276,11 @@ rs9_of_clk_get(struct of_phandle_args *clkspec, void *data)
 	struct rs9_driver_data *rs9 = data;
 	unsigned int idx = clkspec->args[0];
 
+	if (idx >= rs9->chip_info->num_clks) {
+		pr_err("%s: Invalid clock index %u\n", __func__, idx);
+		return ERR_PTR(-EINVAL);
+	}
+
 	return rs9->clk_dif[idx];
 }
 
Signature · 1 line
2.43.0
#2 lore raw

Re: [PATCH v5 1/2] clk: rs9: Add clock index range check to rs9_of_clk_get()

Thank you for your contribution! Sashiko AI review found 3 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 3 potential issue(s) to consider:

Pre-existing issues:
- [High] Hardware configuration is applied after exposing the clock provider to consumers.
- [High] Stack-allocated buffers are used for I2C transfers, violating DMA safety requirements.
- [High] Probe fails entirely if an optional Device Tree property is missing.
--

commit 1a5bae9844abc7b4414cda261d8a9d3cca8fb310
Author: Geert Uytterhoeven <[email protected]>

clk: rs9: Add clock index range check to rs9_of_clk_get()

rs9_of_clk_get() does not validate the clock index in the passed DT clock specifier. If DT specifies an incorrect and out-of-range index, this will access memory beyond the end of the clk_dif[] array.

Fix by this adding a range check to rs9_of_clk_get().

Quoted 4 lines
drivers/clk/clk-renesas-pcie.c +0 -0
index 564550c803896..f3376c29007aa 100644

[ ... ]

drivers/clk/clk-renesas-pcie.c rs9_of_clk_get(struct of_phandle_args *clkspec, void *data) · line 285

Quoted 11 lines
drivers/clk/clk-renesas-pcie.c +5 -0
@@ -276,6 +276,11 @@ rs9_of_clk_get(struct of_phandle_args *clkspec, void *data)
 	struct rs9_driver_data *rs9 = data;
 	unsigned int idx = clkspec->args[0];
 
+	if (idx >= rs9->chip_info->num_clks) {
+		pr_err("%s: Invalid clock index %u\n", __func__, idx);
+		return ERR_PTR(-EINVAL);
+	}
+
 	return rs9->clk_dif[idx];
 }

While reviewing this patch, I noticed a few pre-existing issues in the same file.

drivers/clk/clk-renesas-pcie.c:rs9_probe() {
    ...
	ret = devm_of_clk_add_hw_provider(&client->dev, of_clk_hw_onecell_get,
					  &rs9->onecell);
	if (!ret)
		rs9_update_config(rs9);
    ...
}

[Severity: High]
This isn't a bug introduced by this patch, but does the hardware configuration happen too late during probe?

Since rs9_probe() registers the clock provider via devm_of_clk_add_hw_provider(), exposing it to the system, could a concurrently probing consumer acquire and use the unconfigured clock before rs9_update_config() applies the hardware parameters?

drivers/clk/clk-renesas-pcie.c:rs9_regmap_i2c_write() {
    ...
	const u8 data[3] = { reg, 1, val };
	const int count = ARRAY_SIZE(data);
	int ret;

	ret = i2c_master_send(i2c, data, count);
    ...
}

And similarly in rs9_regmap_i2c_read():

drivers/clk/clk-renesas-pcie.c:rs9_regmap_i2c_read() {
    ...
	u8 txdata = reg;
	u8 rxdata[2];

	...
	xfer[0].buf = (void *)&txdata;

	...
	xfer[1].buf = (void *)rxdata;

	ret = i2c_transfer(i2c->adapter, xfer, 2);
    ...
}

[Severity: High]
This isn't a bug introduced by this patch, but do the custom I2C read and write functions pass stack-allocated buffers to the I2C core?

If the underlying I2C adapter relies on DMA mapping, could passing these stack-allocated buffers (data, txdata, rxdata) to i2c_master_send() and i2c_transfer() trigger a panic or silent memory corruption on architectures with VMAP_STACK enabled?

drivers/clk/clk-renesas-pcie.c:rs9_get_output_config() {
    ...
	ret = of_property_read_u32(np, "renesas,slew-rate", &sr);
	of_node_put(np);
	if (!ret) {
		...
	}

	return ret;
}

[Severity: High]
This isn't a bug introduced by this patch, but can missing an optional Device Tree property cause the probe to fail entirely?

If the optional "renesas,slew-rate" property is absent from a child node, of_property_read_u32() returns -EINVAL. Could this unconditionally propagate back through rs9_probe() and break the clock generator completely, instead of safely falling back to default settings?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
#3 lore raw

PATCH v5 2/2 clk: rs9: Convert to clk_hw_onecell_data and of_clk_hw_onecell_get()

In reply to Geert Uytterhoeven

Convert the rs9 clock driver to use the of_clk_hw_onecell_get() helper, which requires using the clk_hw_onecell_data structure. Embedding that structure in the rs9_driver_data structure has the benefit that the clock array always has the correct size, and thus can no longer become out of sync when adding support for new rs9 variants.

---
v5:
- No changes,

v4:
- Add Reviewed-by,
- Rebase on top of "clk: rs9: Add clock index range check to rs9_of_clk_get()", so this becomes a pure refactoring instead of a fix,
- Put in a series,

(v3 and v2 do not exist)

v1:
- https://lore.kernel.org/a6dce17b15d29a257d09fe0edc199a14c297f1a8.1768836042.git.geert+renesas@glider.be

---
 drivers/clk/clk-renesas-pcie.c | 44 ++++++++++++++--------------------
 1 file changed, 18 insertions(+), 26 deletions(-)
drivers/clk/clk-renesas-pcie.c +18 -26
index f3376c29007aa92d..e2b441f99ccab4c2 100644
@@ -63,10 +63,11 @@ struct rs9_driver_data {
 	struct i2c_client	*client;
 	struct regmap		*regmap;
 	const struct rs9_chip_info *chip_info;
-	struct clk_hw		*clk_dif[8];
 	u8			pll_amplitude;
 	u8			pll_ssc;
 	u8			clk_dif_sr;
+	/* must be last */
+	struct clk_hw_onecell_data onecell;
 };
 
 /*
@@ -270,37 +271,28 @@ static void rs9_update_config(struct rs9_driver_data *rs9)
 	}
 }
 
-static struct clk_hw *
-rs9_of_clk_get(struct of_phandle_args *clkspec, void *data)
-{
-	struct rs9_driver_data *rs9 = data;
-	unsigned int idx = clkspec->args[0];
-
-	if (idx >= rs9->chip_info->num_clks) {
-		pr_err("%s: Invalid clock index %u\n", __func__, idx);
-		return ERR_PTR(-EINVAL);
-	}
-
-	return rs9->clk_dif[idx];
-}
-
 static int rs9_probe(struct i2c_client *client)
 {
+	const struct rs9_chip_info *chip_info;
 	unsigned char name[5] = "DIF0";
 	struct rs9_driver_data *rs9;
 	unsigned int vid, did;
 	struct clk_hw *hw;
 	int i, ret;
 
-	rs9 = devm_kzalloc(&client->dev, sizeof(*rs9), GFP_KERNEL);
+	chip_info = i2c_get_match_data(client);
+	if (!chip_info)
+		return -EINVAL;
+
+	rs9 = devm_kzalloc(&client->dev, struct_size(rs9, onecell.hws,
+			   chip_info->num_clks), GFP_KERNEL);
 	if (!rs9)
 		return -ENOMEM;
 
 	i2c_set_clientdata(client, rs9);
 	rs9->client = client;
-	rs9->chip_info = i2c_get_match_data(client);
-	if (!rs9->chip_info)
-		return -EINVAL;
+	rs9->chip_info = chip_info;
+	rs9->onecell.num = chip_info->num_clks;
 
 	/* Fetch common configuration from DT (if specified) */
 	ret = rs9_get_common_config(rs9);
@@ -308,7 +300,7 @@ static int rs9_probe(struct i2c_client *client)
 		return ret;
 
 	/* Fetch DIFx output configuration from DT (if specified) */
-	for (i = 0; i < rs9->chip_info->num_clks; i++) {
+	for (i = 0; i < rs9->onecell.num; i++) {
 		ret = rs9_get_output_config(rs9, i);
 		if (ret)
 			return ret;
@@ -334,24 +326,24 @@ static int rs9_probe(struct i2c_client *client)
 		return ret;
 
 	vid &= RS9_REG_VID_MASK;
-	if (vid != RS9_REG_VID_IDT || did != rs9->chip_info->did)
+	if (vid != RS9_REG_VID_IDT || did != chip_info->did)
 		return dev_err_probe(&client->dev, -ENODEV,
 				     "Incorrect VID/DID: %#02x, %#02x. Expected %#02x, %#02x\n",
-				     vid, did, RS9_REG_VID_IDT,
-				     rs9->chip_info->did);
+				     vid, did, RS9_REG_VID_IDT, chip_info->did);
 
 	/* Register clock */
-	for (i = 0; i < rs9->chip_info->num_clks; i++) {
+	for (i = 0; i < rs9->onecell.num; i++) {
 		snprintf(name, 5, "DIF%d", i);
 		hw = devm_clk_hw_register_fixed_factor_index(&client->dev, name,
 						    0, 0, 4, 1);
 		if (IS_ERR(hw))
 			return PTR_ERR(hw);
 
-		rs9->clk_dif[i] = hw;
+		rs9->onecell.hws[i] = hw;
 	}
 
-	ret = devm_of_clk_add_hw_provider(&client->dev, rs9_of_clk_get, rs9);
+	ret = devm_of_clk_add_hw_provider(&client->dev, of_clk_hw_onecell_get,
+					  &rs9->onecell);
 	if (!ret)
 		rs9_update_config(rs9);
 
Signature · 1 line
2.43.0
Applied
#4 lore raw
In reply to Geert Uytterhoeven
Quoted 10 lines · Geert Uytterhoeven

clk: rs9: Fix and of_clk_hw_onecell_get() conversion

Hi all,

This patch series contains a fix and further development for the Renesas 9-series PCIe clock generator driver, which were submitted before as separate patches As the fix hasn't been picked up yet, and the further development depends on the fix, I am resending them as a series.

[...]

Applied, thanks!

[1/2] clk: rs9: Add clock index range check to rs9_of_clk_get()

      commit: a2ef401bdfd19aa41d0e3e1b630a12fbc01ac184
[2/2] clk: rs9: Convert to clk_hw_onecell_data and of_clk_hw_onecell_get()
      commit: 07f4bac13786a80cd3e6547051376ea65da2f4c5

Best regards,

Signature · 1 line
Brian Masney <[email protected]>