Hi Robin,
> Am 25.10.2024 um 16:27 schrieb Robin Murphy <robin.murphy@xxxxxxx>:
>
> On 08/10/2024 8:52 pm, H. Nikolaus Schaller wrote:
>>> Am 08.10.2024 um 20:20 schrieb Robin Murphy <robin.murphy@xxxxxxx>:
>>>> The -ETIMEDOUT seems to come from of_iommu_configure().
>>>
>
> Hmm, I dug around and found a Pandaboard in the cupboard, and ostensibly this seems to work as expected there:
>
> 6:12-rc3:
> chu-chu-rocket:~ # dmesg | grep -i iommu
> [ 0.628601] iommu: Default domain type: Translated
> [ 0.633575] iommu: DMA domain TLB invalidation policy: strict mode
> [ 1.636047] omap-iommu 4a066000.mmu: 4a066000.mmu registered
> [ 3.265869] omap-iommu 55082000.mmu: 55082000.mmu registered
>
> 6.12-rc3 + of_xlate:
> chu-chu-rocket:~ # dmesg | grep -i iommu
> [ 0.629577] iommu: Default domain type: Translated
> [ 0.634582] iommu: DMA domain TLB invalidation policy: strict mode
> [ 1.622802] omap-iommu 4a066000.mmu: 4a066000.mmu registered
> [ 3.316009] omap-iommu 55082000.mmu: 55082000.mmu registered
> [ 3.329040] omap-rproc ocp:dsp: Adding to iommu group 0
> [ 3.335083] omap-iommu 4a066000.mmu: 4a066000.mmu: version 2.0
> [ 3.356506] omap-rproc 55020000.ipu: Adding to iommu group 1
> [ 3.362396] omap-iommu 55082000.mmu: 55082000.mmu: version 2.1
>
> Guess I'm going to have to dig further for an OMAP3 to figure out what additional shenanigans that ISP driver is up to... :/
I was able to confirm on 6.12-rc4 ecept a small difference in the last lines (ocp:dsp and no version 2.x):
root@letux:~# dmesg|fgrep iommu
[ 0.402191] iommu: Default domain type: Translated
[ 0.402221] iommu: DMA domain TLB invalidation policy: strict mode
[ 0.741607] omap-iommu 4a066000.mmu: 4a066000.mmu registered
[ 2.461120] omap-iommu 55082000.mmu: 55082000.mmu registered
[ 2.474914] platform ocp:dsp: Adding to iommu group 0
[ 2.481719] platform 55020000.ipu: Adding to iommu group 1
root@letux:~#
On OMAP3 it is similar. The patch is doing something.
I get this w/o patch:
root@letux:~# dmesg|fgrep iommu
[ 0.515808] iommu: Default domain type: Translated
[ 0.515808] iommu: DMA domain TLB invalidation policy: strict mode
[ 0.528686] omap-iommu 480bd400.mmu: 480bd400.mmu registered
root@letux:~# dmesg|fgrep .isp
[ 10.924438] omap3isp 480bc000.isp: supply vdd-csiphy1 not found, using dummy regulator
[ 10.989044] omap3isp 480bc000.isp: supply vdd-csiphy2 not found, using dummy regulator
[ 11.082916] omap3isp 480bc000.isp: Revision 15.0 found
[ 11.107879] omap3isp 480bc000.isp: failed to create ARM IOMMU mapping
[ 11.160736] omap3isp 480bc000.isp: unable to attach to IOMMU
[ 11.181640] omap3isp 480bc000.isp: isp_xclk_set_rate: cam_xclka set to 24685714 Hz (div 7)
root@letux:~#
with patch:
root@letux:~# dmesg|fgrep iommu
[ 0.522552] iommu: Default domain type: Translated
[ 0.522583] iommu: DMA domain TLB invalidation policy: strict mode
[ 0.525543] platform 480bc000.isp: Adding to iommu group 0
[ 0.525726] omap-iommu 480bd400.mmu: 480bd400.mmu registered
[ 34.422973] omap-iommu 480bd400.mmu: 480bd400.mmu: version 1.1
root@letux:~# dmesg|fgrep .isp
[ 0.525543] platform 480bc000.isp: Adding to iommu group 0
[ 34.413330] omap3isp 480bc000.isp: deferred probe timeout, ignoring dependency
[ 34.429412] omap3isp 480bc000.isp: supply vdd-csiphy1 not found, using dummy regulator
[ 34.446655] omap3isp 480bc000.isp: supply vdd-csiphy2 not found, using dummy regulator
[ 34.456115] omap3isp 480bc000.isp: Revision 15.0 found
[ 34.463165] omap3isp 480bc000.isp: failed to attach device to VA mapping
[ 34.470306] omap3isp 480bc000.isp: unable to attach to IOMMU
[ 34.477661] omap3isp 480bc000.isp: probe with driver omap3isp failed with error -16
root@letux:~#
Maybe I am missing something from your latest patches. I currently have:
diff --git a/drivers/iommu/omap-iommu.c b/drivers/iommu/omap-iommu.c
index c9528065a59af..de14a3cd4b426 100644
--- a/drivers/iommu/omap-iommu.c
+++ b/drivers/iommu/omap-iommu.c
@@ -1230,7 +1230,8 @@ static int omap_iommu_probe(struct platform_device *pdev)
if (err)
return err;
- err = iommu_device_register(&obj->iommu, &omap_iommu_ops, &pdev->dev);
+ err = iommu_device_register(&obj->iommu, &omap_iommu_ops, NULL);
if (err)
goto out_sysfs;
obj->has_iommu_driver = true;
@@ -1691,7 +1692,8 @@ static struct iommu_device *omap_iommu_probe_device(struct device *dev)
if (!oiommu) {
of_node_put(np);
kfree(arch_data);
- return ERR_PTR(-EINVAL);
+ /* Not fatal, will re-probe once the right instance registers itself */
+ return ERR_PTR(-ENODEV);
}
tmp->iommu_dev = oiommu;
Do you have a link to the patches you have tested?
What I have also compared are the DTS and DTSI files of omap3/36xx and omap4/omap4-l4.
Although there are architectural differences I didn't find anything significant.
More different is how the omap3-isp vs. omap4-iva is handled.
omap3-isp is a media driver calling arm_iommu_create_mapping() which fails.
omap4-iva is a remoteproc driver using a different strategy (omap_rproc_of_get_internal_memories ?)
where I do not know details where iommu comes into the game.
Hope this helps.
Best regards,
Nikolaus
