diff mbox series

[v2,04/11] thermal: k3_j72xx_bandgap: map fuse_base only for erratum workaround

Message ID 20221025191515.9151-5-bb@ti.com (mailing list archive)
State New, archived
Headers show
Series enable VTM node for all TI's K3 SoCs | expand

Commit Message

Bryan Brattlof Oct. 25, 2022, 7:15 p.m. UTC 
Some of TI's J721E SoCs require a software trimming procedure for the
temperature monitors to function properly. To determine if a particular
J721E is not affected by this erratum, both bits in the WKUP_SPARE_FUSE0
region must be set. Other SoCs, not affected by this erratum, will not
need this region.

Map the 'fuse_base' region only when the erratum fix is needed.

Signed-off-by: Bryan Brattlof <bb@ti.com>
---
 drivers/thermal/k3_j72xx_bandgap.c | 37 ++++++++++++++++++------------
 1 file changed, 22 insertions(+), 15 deletions(-)

Comments

Bryan Brattlof Oct. 25, 2022, 7:42 p.m. UTC | #1 
On October 25, 2022 thus sayeth Bryan Brattlof:
> Some of TI's J721E SoCs require a software trimming procedure for the
> temperature monitors to function properly. To determine if a particular
> J721E is not affected by this erratum, both bits in the WKUP_SPARE_FUSE0
> region must be set. Other SoCs, not affected by this erratum, will not
> need this region.
> 
> Map the 'fuse_base' region only when the erratum fix is needed.
> 
> Signed-off-by: Bryan Brattlof <bb@ti.com>
> ---
>  drivers/thermal/k3_j72xx_bandgap.c | 37 ++++++++++++++++++------------
>  1 file changed, 22 insertions(+), 15 deletions(-)
> 

...

>  
> -	if (workaround_needed)
> -		devm_iounmap(dev, fuse_base);
> -

Ugh... This is wrong. 

Of course I will catch it right after I send this out

Sorry for the noise everyone
~Bryan
diff mbox series

Patch

diff --git a/drivers/thermal/k3_j72xx_bandgap.c b/drivers/thermal/k3_j72xx_bandgap.c
index 71bf65e6afaed..031ea1091909a 100644
--- a/drivers/thermal/k3_j72xx_bandgap.c
+++ b/drivers/thermal/k3_j72xx_bandgap.c
@@ -386,15 +386,32 @@  static int k3_j72xx_bandgap_probe(struct platform_device *pdev)
 	if (IS_ERR(bgp->cfg2_base))
 		return PTR_ERR(bgp->cfg2_base);
 
-	res = platform_get_resource(pdev, IORESOURCE_MEM, 2);
-	fuse_base = devm_ioremap_resource(dev, res);
-	if (IS_ERR(fuse_base))
-		return PTR_ERR(fuse_base);
-
 	driver_data = of_device_get_match_data(dev);
 	if (driver_data)
 		workaround_needed = driver_data->has_errata_i2128;
 
+	/*
+	 * Some of TI's J721E SoCs require a software trimming procedure
+	 * for the temperature monitors to function properly. To determine
+	 * if this particular SoC is NOT affected, both bits in the
+	 * WKUP_SPARE_FUSE0[31:30] will be set (0xC0000000) indicating
+	 * when software trimming should NOT be applied.
+	 *
+	 * https://www.ti.com/lit/er/sprz455c/sprz455c.pdf
+	 */
+	if (workaround_needed) {
+		res = platform_get_resource(pdev, IORESOURCE_MEM, 2);
+		fuse_base = devm_ioremap_resource(dev, res);
+		if (IS_ERR(fuse_base))
+			return PTR_ERR(fuse_base);
+
+		if ((readl(fuse_base) & 0xc0000000) == 0xc0000000)
+			workaround_needed = false;
+	}
+
+	dev_dbg(bgp->dev, "Work around %sneeded\n",
+		workaround_needed ? "" : "not ");
+
 	pm_runtime_enable(dev);
 	ret = pm_runtime_get_sync(dev);
 	if (ret < 0) {
@@ -427,13 +444,6 @@  static int k3_j72xx_bandgap_probe(struct platform_device *pdev)
 		goto err_free_ref_table;
 	}
 
-	/* Workaround not needed if bit30/bit31 is set even for J721e */
-	if (workaround_needed && (readl(fuse_base + 0x0) & 0xc0000000) == 0xc0000000)
-		workaround_needed = false;
-
-	dev_dbg(bgp->dev, "Work around %sneeded\n",
-		workaround_needed ? "" : "not ");
-
 	if (!workaround_needed)
 		init_table(5, ref_table, golden_factors);
 	else
@@ -501,9 +511,6 @@  static int k3_j72xx_bandgap_probe(struct platform_device *pdev)
 	 */
 	kfree(ref_table);
 
-	if (workaround_needed)
-		devm_iounmap(dev, fuse_base);
-
 	return 0;
 
 err_free_ref_table: