> -----Original Message-----
> From: Greg KH [mailto:gre...@linuxfoundation.org]
> Sent: Tuesday, June 2, 2020 6:11 PM
> To: Song Bao Hua (Barry Song) <song.bao....@hisilicon.com>
> Cc: raf...@kernel.org; io...@lists.linux-foundation.org;
> linux-arm-ker...@lists.infradead.org; linux-kernel@vger.kernel.org; Linuxarm
> <linux...@huawei.com>; Zengtao (B) <prime.z...@hisilicon.com>; Robin
> Murphy <robin.mur...@arm.com>
> Subject: Re: [PATCH] driver core: platform: expose numa_node to users in sysfs
>
> On Tue, Jun 02, 2020 at 05:09:57AM +0000, Song Bao Hua (Barry Song) wrote:
> > > >
> > > > Platform devices are NUMA? That's crazy, and feels like a total
> > > > abuse of platform devices and drivers that really should belong on a
> "real"
> > > > bus.
> > >
> > > I am not sure if it is an abuse of platform device. But smmu is a
> > > platform device, drivers/iommu/arm-smmu-v3.c is a platform driver.
> > > In a typical ARM server, there are maybe multiple SMMU devices which
> > > can support IO virtual address and page tables for other devices on
> > > PCI-like busses.
> > > Each different SMMU device might be close to different NUMA node.
> > > There is really a hardware topology.
> > >
> > > If you have multiple CPU packages in a NUMA server, some platform
> > > devices might Belong to CPU0, some other might belong to CPU1.
> >
> > Those devices are populated by acpi_iort for an ARM server:
> >
> > drivers/acpi/arm64/iort.c:
> >
> > static const struct iort_dev_config iort_arm_smmu_v3_cfg __initconst = {
> > .name = "arm-smmu-v3",
> > .dev_dma_configure = arm_smmu_v3_dma_configure,
> > .dev_count_resources = arm_smmu_v3_count_resources,
> > .dev_init_resources = arm_smmu_v3_init_resources,
> > .dev_set_proximity = arm_smmu_v3_set_proximity, };
> >
> > void __init acpi_iort_init(void)
> > {
> > acpi_status status;
> >
> > status = acpi_get_table(ACPI_SIG_IORT, 0, &iort_table);
> > ...
> > iort_check_id_count_workaround(iort_table);
> > iort_init_platform_devices();
> > }
> >
> > static void __init iort_init_platform_devices(void) {
> > ...
> >
> > for (i = 0; i < iort->node_count; i++) {
> > if (iort_node >= iort_end) {
> > pr_err("iort node pointer overflows, bad
> table\n");
> > return;
> > }
> >
> > iort_enable_acs(iort_node);
> >
> > ops = iort_get_dev_cfg(iort_node);
> > if (ops) {
> > fwnode = acpi_alloc_fwnode_static();
> > if (!fwnode)
> > return;
> >
> > iort_set_fwnode(iort_node, fwnode);
> >
> > ret = iort_add_platform_device(iort_node, ops);
> > if (ret) {
> > iort_delete_fwnode(iort_node);
> > acpi_free_fwnode_static(fwnode);
> > return;
> > }
> > }
> >
> > ...
> > }
> > ...
> > }
> >
> > NUMA node is got from ACPI:
> >
> > static int __init arm_smmu_v3_set_proximity(struct device *dev,
> > struct acpi_iort_node
> > *node) {
> > struct acpi_iort_smmu_v3 *smmu;
> >
> > smmu = (struct acpi_iort_smmu_v3 *)node->node_data;
> > if (smmu->flags & ACPI_IORT_SMMU_V3_PXM_VALID) {
> > int dev_node = acpi_map_pxm_to_node(smmu->pxm);
> >
> > ...
> >
> > set_dev_node(dev, dev_node);
> > ...
> > }
> > return 0;
> > }
> >
> > Barry
>
> That's fine, but those are "real" devices, not platform devices, right?
>
Most platform devices are "real" memory-mapped hardware devices. For an
embedded system, almost all "simple-bus"
devices are populated from device trees as platform devices. Only a part of
platform devices are not "real" hardware.
Smmu is a memory-mapped device. It is totally like most other platform devices
populated in a
memory space mapped in cpu's local space. It uses ioremap to map registers, use
readl/writel to read/write its
space.
> What platform device has this issue? What one will show up this way with
> the new patch?
if platform device shouldn't be a real hardware, there is no platform device
with a hardware topology.
But platform devices are "real" hardware at most time. Smmu is a "real" device,
but it is a platform device in Linux.
>
> thanks,
>
> greg k-h
-barry
