On Wed, Mar 24, 2021 at 09:15:46AM +0000, Gustavo Pimentel wrote:
> Hi Greg,
>
> On Tue, Mar 23, 2021 at 13:0:4, Greg Kroah-Hartman
> <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
>
> > On Fri, Feb 12, 2021 at 06:28:03PM +0100, Gustavo Pimentel wrote:
> > > +static const struct attribute_group xdata_attr_group = {
> > > + .attrs = default_attrs,
> > > + .name = DW_XDATA_DRIVER_NAME,
> > > +};
> >
> > ATTRIBUTE_GROUPS()?
>
> Nicely catched!
>
> >
> > > +static int dw_xdata_pcie_probe(struct pci_dev *pdev,
> > > + const struct pci_device_id *pid)
> > > +{
> > > + const struct dw_xdata_pcie_data *pdata = (void *)pid->driver_data;
> > > + struct dw_xdata *dw;
> > > + u64 addr;
> > > + int err;
> > > +
> > > + /* Enable PCI device */
> > > + err = pcim_enable_device(pdev);
> > > + if (err) {
> > > + pci_err(pdev, "enabling device failed\n");
> > > + return err;
> > > + }
> > > +
> > > + /* Mapping PCI BAR regions */
> > > + err = pcim_iomap_regions(pdev, BIT(pdata->rg_bar), pci_name(pdev));
> > > + if (err) {
> > > + pci_err(pdev, "xData BAR I/O remapping failed\n");
> > > + return err;
> > > + }
> > > +
> > > + pci_set_master(pdev);
> > > +
> > > + /* Allocate memory */
> > > + dw = devm_kzalloc(&pdev->dev, sizeof(*dw), GFP_KERNEL);
> > > + if (!dw)
> > > + return -ENOMEM;
> > > +
> > > + /* Data structure initialization */
> > > + mutex_init(&dw->mutex);
> > > +
> > > + dw->rg_region.vaddr = pcim_iomap_table(pdev)[pdata->rg_bar];
> > > + if (!dw->rg_region.vaddr)
> > > + return -ENOMEM;
> > > +
> > > + dw->rg_region.vaddr += pdata->rg_off;
> > > + dw->rg_region.paddr = pdev->resource[pdata->rg_bar].start;
> > > + dw->rg_region.paddr += pdata->rg_off;
> > > + dw->rg_region.sz = pdata->rg_sz;
> > > +
> > > + dw->max_wr_len = pcie_get_mps(pdev);
> > > + dw->max_wr_len >>= 2;
> > > +
> > > + dw->max_rd_len = pcie_get_readrq(pdev);
> > > + dw->max_rd_len >>= 2;
> > > +
> > > + dw->pdev = pdev;
> > > +
> > > + writel(0x0, &(__dw_regs(dw)->RAM_addr));
> > > + writel(0x0, &(__dw_regs(dw)->RAM_port));
> > > +
> > > + addr = dw->rg_region.paddr + DW_XDATA_EP_MEM_OFFSET;
> > > + writel(lower_32_bits(addr), &(__dw_regs(dw)->addr_lsb));
> > > + writel(upper_32_bits(addr), &(__dw_regs(dw)->addr_msb));
> > > + pci_dbg(pdev, "xData: target address = 0x%.16llx\n", addr);
> > > +
> > > + pci_dbg(pdev, "xData: wr_len=%zu, rd_len=%zu\n",
> > > + dw->max_wr_len * 4, dw->max_rd_len * 4);
> > > +
> > > + /* Saving data structure reference */
> > > + pci_set_drvdata(pdev, dw);
> > > +
> > > + /* Sysfs */
> > > + err = sysfs_create_group(&pdev->dev.kobj, &xdata_attr_group);
> >
> > You just raced with userspace and lost :(
> >
> > Have the driver core properly create/remove your sysfs files, set the
> > default groups pointer in your driver and all will be fine.
>
> I've gone around and around, searched in other drivers, but I'm not
> understanding your PoV or what I should do. Can you throw me a bone here?
> I'm starting to pull my hair off, lol
Set the dev_groups field of the struct device_driver in your pci driver
structure and all should be fine.
But wait, why are you adding attributes to a pci device and not your
"own" device? You should make a sub-device for this device-specific
things, otherwise it will be impossible to find these types of
attributes anywhere in sysfs.
> I would like to ask you about something, it's you the maintainer of misc
> drivers right?
MAINTAINERS shows this so I guess it must be true...
> After fixing this issue, does this driver have a chance to be pulled on
> 5.13?
No idea, I do not promise anything :)
thanks,
greg k-h
