linux-kernel - Re: [PATCH v3 03/27] PCI: pci resource iterator

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAErSpo7TReTzz289EMuh5gGKjskCLmACNT99o1s-QCzRHLGaXg@mail.gmail.com>
Date:	Wed, 10 Apr 2013 09:22:48 -0600
From:	Bjorn Helgaas <bhelgaas@...gle.com>
To:	Ram Pai <linuxram@...ibm.com>
Cc:	Yinghai Lu <yinghai@...nel.org>,
	"linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3 03/27] PCI: pci resource iterator

On Mon, Apr 8, 2013 at 10:51 PM, Ram Pai <linuxram@...ibm.com> wrote:
> On Thu, Apr 04, 2013 at 04:18:01PM -0600, Bjorn Helgaas wrote:
>> On Wed, Mar 13, 2013 at 5:27 PM, Yinghai Lu <yinghai@...nel.org> wrote:
>> > From: Ram Pai <linuxram@...ibm.com>
>> >
>> > Currently pci_dev structure holds an array of 17 PCI resources; six base
>> > BARs, one ROM BAR, four BRIDGE BARs, six sriov BARs.  This is wasteful.
>> > A bridge device just needs the 4 bridge resources. A non-bridge device
>> > just needs the six base resources and one ROM resource. The sriov
>> > resources are needed only if the device has SRIOV capability.
>> >
>> > The pci_dev structure needs to be re-organized to avoid unnecessary
>> > bloating.  However too much code outside the pci-bus driver, assumes the
>> > internal details of the pci_dev structure, thus making it hard to
>> > re-organize the datastructure.
>> >
>> > As a first step this patch provides generic methods to access the
>> > resource structure of the pci_dev.
>> >
>> > Finally we can re-organize the resource structure in the pci_dev
>> > structure and correspondingly update the methods.
>> >
>> > -v2: Consolidated iterator interface as per Bjorn's suggestion.
>> > -v3: Add the idx back - Yinghai Lu
>> > -v7: Change to use bitmap for searching - Yinghai Lu
>> > -v8: Fix acpiphp module compiling error that is found by
>> >         Steven Newbury <steve@...wbury.org.uk> - Yinghai Lu
>> >
>> > Signed-off-by: Ram Pai <linuxram@...ibm.com>
>> > Signed-off-by: Yinghai Lu <yinghai@...nel.org>
>> > ---
>> >  drivers/pci/probe.c |   48 ++++++++++++++++++++++++++++++++++++++++++++++++
>> >  include/linux/pci.h |   24 ++++++++++++++++++++++++
>> >  2 files changed, 72 insertions(+)
>> >
>> > diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
>> > index 1df75f7..ac751a6 100644
>> > --- a/drivers/pci/probe.c
>> > +++ b/drivers/pci/probe.c
>> > @@ -123,6 +123,54 @@ int pci_dev_resource_idx(struct pci_dev *dev, struct resource *res)
>> >         return -1;
>> >  }
>> >
>> > +static void __init_res_idx_mask(unsigned long *mask, int flag)
>> > +{
>> > +       bitmap_zero(mask, PCI_NUM_RESOURCES);
>> > +       if (flag & PCI_STD_RES)
>> > +               bitmap_set(mask, PCI_STD_RESOURCES,
>> > +                       PCI_STD_RESOURCE_END - PCI_STD_RESOURCES + 1);
>> > +       if (flag & PCI_ROM_RES)
>> > +               bitmap_set(mask, PCI_ROM_RESOURCE, 1);
>> > +#ifdef CONFIG_PCI_IOV
>> > +       if (flag & PCI_IOV_RES)
>> > +               bitmap_set(mask, PCI_IOV_RESOURCES,
>> > +                       PCI_IOV_RESOURCE_END - PCI_IOV_RESOURCES + 1);
>> > +#endif
>> > +       if (flag & PCI_BRIDGE_RES)
>> > +               bitmap_set(mask, PCI_BRIDGE_RESOURCES,
>> > +                       PCI_BRIDGE_RESOURCE_END - PCI_BRIDGE_RESOURCES + 1);
>> > +}
>> > +
>> > +static DECLARE_BITMAP(res_idx_mask[1 << PCI_RES_BLOCK_NUM], PCI_NUM_RESOURCES);
>> > +static int __init pci_res_idx_mask_init(void)
>> > +{
>> > +       int i;
>> > +
>> > +       for (i = 0; i < (1 << PCI_RES_BLOCK_NUM); i++)
>> > +               __init_res_idx_mask(res_idx_mask[i], i);
>> > +
>> > +       return 0;
>> > +}
>> > +postcore_initcall(pci_res_idx_mask_init);
>> > +
>> > +static inline unsigned long *get_res_idx_mask(int flag)
>> > +{
>> > +       return res_idx_mask[flag & ((1 << PCI_RES_BLOCK_NUM) - 1)];
>> > +}
>> > +
>> > +int pci_next_resource_idx(int i, int flag)
>> > +{
>> > +       i++;
>> > +       if (i < PCI_NUM_RESOURCES)
>> > +               i = find_next_bit(get_res_idx_mask(flag), PCI_NUM_RESOURCES, i);
>> > +
>> > +       if (i < PCI_NUM_RESOURCES)
>> > +               return i;
>> > +
>> > +       return -1;
>> > +}
>> > +EXPORT_SYMBOL(pci_next_resource_idx);
>> > +
>> >  static u64 pci_size(u64 base, u64 maxbase, u64 mask)
>> >  {
>> >         u64 size = mask & maxbase;      /* Find the significant bits */
>> > diff --git a/include/linux/pci.h b/include/linux/pci.h
>> > index aefff8b..127a856 100644
>> > --- a/include/linux/pci.h
>> > +++ b/include/linux/pci.h
>> > @@ -341,6 +341,30 @@ struct pci_dev {
>> >  struct resource *pci_dev_resource_n(struct pci_dev *dev, int n);
>> >  int pci_dev_resource_idx(struct pci_dev *dev, struct resource *res);
>> >
>> > +#define PCI_STD_RES            (1<<0)
>> > +#define PCI_ROM_RES            (1<<1)
>> > +#define PCI_IOV_RES            (1<<2)
>> > +#define PCI_BRIDGE_RES         (1<<3)
>> > +#define PCI_RES_BLOCK_NUM      4
>> > +
>> > +#define PCI_ALL_RES            (PCI_STD_RES | PCI_ROM_RES | PCI_BRIDGE_RES | PCI_IOV_RES)
>> > +#define PCI_NOSTD_RES          (PCI_ALL_RES & ~PCI_STD_RES)
>> > +#define PCI_NOIOV_RES          (PCI_ALL_RES & ~PCI_IOV_RES)
>> > +#define PCI_NOROM_RES          (PCI_ALL_RES & ~PCI_ROM_RES)
>> > +#define PCI_NOBRIDGE_RES       (PCI_ALL_RES & ~PCI_BRIDGE_RES)
>> > +#define PCI_STD_ROM_RES                (PCI_STD_RES | PCI_ROM_RES)
>> > +#define PCI_STD_IOV_RES                (PCI_STD_RES | PCI_IOV_RES)
>> > +#define PCI_STD_ROM_IOV_RES    (PCI_STD_RES | PCI_ROM_RES | PCI_IOV_RES)
>> > +
>> > +int pci_next_resource_idx(int i, int flag);
>> > +
>> > +#define for_each_pci_resource(dev, res, i, flag)       \
>> > +       for (i = pci_next_resource_idx(-1, flag),       \
>> > +               res = pci_dev_resource_n(dev, i);       \
>> > +            res;                                       \
>> > +            i = pci_next_resource_idx(i, flag),        \
>> > +               res = pci_dev_resource_n(dev, i))
>> > +
>> >  static inline struct pci_dev *pci_physfn(struct pci_dev *dev)
>> >  {
>> >  #ifdef CONFIG_PCI_IOV
>>
>> I think this turned out to be a disaster, with all the bitmaps and
>> helper functions.  Filtering in the bodies of the
>> for_each_pci_resource() users has *got* to be better.  That probably
>> requires a wrapper struct around the struct resource.  Or possibly
>> having a wrapper struct with a "type" or "flags" field would make
>> filtering in for_each_pci_resources() itself cleaner to implement.
>
> I agree. There are two cleanups needed.
>
> a) pci drivers should not assume the internal organization of the
>         resources in the struct pci_dev.

Do you mean that drivers should not use "pci_dev->resource[i]"?  If
so, I agree that it would be great if we had an accessor for BARs, but
it seems impractical to change all the drivers that use the current
style.

The number of places that actually look at *non-BAR* pci_dev
resources, e.g., the places that look at bridge windows and SR-IOV
BARs, should be pretty small, and it seems reasonable to change them.

> b) The type of a resource has to be determined based on some
>    information internal to the resource; possibly a flag,
>    instead of the relative position of the resource in some array.

Yes, I agree.

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