Re: [PATCH] PCI: save/restore max Latency Value for device LTR

From: Bjorn Helgaas
Date: Fri May 18 2012 - 20:27:45 EST


On Tue, May 8, 2012 at 3:09 AM, Hao, Xudong <xudong.hao@xxxxxxxxx> wrote:
>> -----Original Message-----
>> From: Bjorn Helgaas [mailto:bhelgaas@xxxxxxxxxx]
>> >  }
>> >
>>
>> This doesn't make any sense to me.  "pos" is the offset of the PCI
>> Express Capability (identifier 10h).  LTR is a separate extended
>> capability (identifier 18h), so you at least have to look up its
>> offset.
>>
> Sorry paste a wrong patch...
>
> How about this patch, not a formal patch.
>
> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
> index 111569c..eced407 100644
> --- a/drivers/pci/pci.c
> +++ b/drivers/pci/pci.c
> @@ -910,6 +910,45 @@ static void pci_restore_pcie_state(struct pci_dev *dev)
>                pci_write_config_word(dev, pos + PCI_EXP_SLTCTL2, cap[i++]);
>  }
>
> +static int pci_save_ltr_value(struct pci_dev *dev)
> +{
> +       int i = 0, pos;
> +       struct pci_cap_saved_state *save_state;
> +       u16 *cap;
> +
> +       pos = pci_find_ext_capability(dev, PCI_EXT_CAP_ID_LTR);
> +       if (!pos)
> +               return -ENOTSUPP;
> +
> +       save_state = pci_find_saved_cap(dev, PCI_EXT_CAP_ID_LTR);
> +       if (!save_state) {
> +               dev_err(&dev->dev, "buffer not found in %s\n", __func__);
> +               return -ENOMEM;
> +       }
> +       cap = (u16 *)&save_state->cap.ltr_data[0];
> +
> +       pci_read_config_word(dev, pos + PCI_LTR_MAX_SNOOP_LAT, &cap[i++]);
> +       pci_read_config_word(dev, pos + PCI_LTR_MAX_NOSNOOP_LAT, &cap[i++]);
> +}
> +
> +static void pci_restore_ltr_value(struct pci_dev *dev)
> +{
> +       int i = 0, pos;
> +       struct pci_cap_saved_state *save_state;
> +       u16 *cap;
> +
> +       if (!pci_ltr_supported(dev))
> +               return;
> +
> +       save_state = pci_find_saved_cap(dev, PCI_EXT_CAP_ID_LTR);
> +       pos = pci_find_ext_capability(dev, PCI_EXT_CAP_ID_LTR);
> +       if (!save_state || !pos)
> +               return;
> +       cap = (u16 *)&save_state->cap.ltr_data[0];
> +
> +       pci_write_config_word(dev, pos + PCI_LTR_MAX_SNOOP_LAT, cap[i++]);
> +       pci_write_config_word(dev, pos + PCI_LTR_MAX_NOSNOOP_LAT, cap[i++]);
> +}
>
>  static int pci_save_pcix_state(struct pci_dev *dev)
>  {
> @@ -964,6 +1003,10 @@ pci_save_state(struct pci_dev *dev)
>                return i;
>        if ((i = pci_save_pcix_state(dev)) != 0)
>                return i;
> +
> +       if (pci_ltr_supported(dev))
> +               return pci_save_ltr_value(dev);
> +
>        return 0;
>  }
>
> @@ -1032,6 +1075,7 @@ void pci_restore_state(struct pci_dev *dev)
>        pci_restore_pcix_state(dev);
>        pci_restore_msi_state(dev);
>        pci_restore_iov_state(dev);
> +       pci_restore_ltr_value(dev);
>
>        dev->state_saved = false;
>  }
> diff --git a/include/linux/pci.h b/include/linux/pci.h
> index e444f5b..6343aeb 100644
> --- a/include/linux/pci.h
> +++ b/include/linux/pci.h
> @@ -220,6 +220,7 @@ struct pci_cap_saved_data {
>        char cap_nr;
>        unsigned int size;
>        u32 data[0];
> +       u32 ltr_data[0];
>  };
>
>  struct pci_cap_saved_state {

Did you test this?

I can't believe this works at all. You're calling
"pci_find_saved_cap(dev, PCI_EXT_CAP_ID_LTR)", which searches the
pci_dev->saved_cap_space list, but you never add a PCI_EXT_CAP_ID_LTR
buffer to that list.

Bjorn
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/