Re: [PATCH -mm] 2/2: PCI: disable decode of IO/memory during BAR sizing

[Jesse Barnes]

On Tuesday, May 29, 2007 9:05:24 Robert Hancock wrote:
> Change PCI BAR sizing to disable the decode of memory or IO, as
> appropriate, while we are writing the all-ones value to the BAR to
> determine the size. If this is not done, the device may spuriously decode
> accesses to memory areas it should not. On some Intel PCI Express chipsets,
> this breaks MMCONFIG configuration space access, since the memory the
> graphics card ends up decoding during this period overlaps the MMCONFIG
> area, and thus it steals the accesses to the area to do any other
> configuration space access, including changing the BAR back to its previous
> value.
>
> However, don't do this disabling on host bridge devices, as it is reported
> that some of them do silly things like disable CPU to RAM access if this is
> done.
>
> Based on an original patch by Jesse Barnes.
>
> Signed-off-by: Robert Hancock <hancockr@xxxxxxx>
>
> --- linux-2.6.22-rc2-mm1/drivers/pci/probe.c	2007-05-23 21:21:05.000000000
> -0600 +++ linux-2.6.22-rc2-mm1edit/drivers/pci/probe.c	2007-05-29
> 21:31:47.000000000 -0600 @@ -180,6 +180,58 @@ static inline int
> is_64bit_memory(u32 ma
>  	return 0;
>  }
>
> +#define BAR_IS_MEMORY(bar) (((bar) & PCI_BASE_ADDRESS_SPACE) ==	\
> +			    PCI_BASE_ADDRESS_SPACE_MEMORY)
> +
> +/**
> + * pci_bar_size - get raw PCI BAR size
> + * @dev: PCI device
> + * @reg: BAR to probe
> + *
> + * Use basic PCI probing:
> + *   - save original BAR value
> + *   - disable MEM or IO decode in PCI_COMMAND reg if appropriate
> + *   - write all 1s to the BAR
> + *   - read back value
> + *   - reenble MEM or IO decode as necessary
> + *   - write original value back
> + *
> + * Returns raw BAR size to caller.
> + */
> +static u32 pci_bar_size(struct pci_dev *dev, unsigned int reg)
> +{
> +	u32 orig_reg, sz;
> +	u16 orig_cmd;
> +
> +	pci_read_config_dword(dev, reg, &orig_reg);
> +	pci_read_config_word(dev, PCI_COMMAND, &orig_cmd);
> +
> +	/*
> +	 * Disable memory or IO decode on the device while writing the test
> +	 * value to the BAR. This prevents possible spurious decoding
> +	 * of random addresses by the device. Don't do this for host bridges,
> +	 * however, since some of them do silly things like disable CPU to RAM
> +	 * access if this is done.
> +	 */
> +	if ((dev->class >> 8) != PCI_CLASS_BRIDGE_HOST) {
> +		if (BAR_IS_MEMORY(orig_reg))
> +			pci_write_config_word(dev, PCI_COMMAND,
> +					      orig_cmd & ~PCI_COMMAND_MEMORY);
> +		else
> +			pci_write_config_word(dev, PCI_COMMAND,
> +					      orig_cmd & ~PCI_COMMAND_IO);
> +	}
> +
> +	pci_write_config_dword(dev, reg, 0xffffffff);
> +	pci_read_config_dword(dev, reg, &sz);
> +	pci_write_config_dword(dev, reg, orig_reg);
> +
> +	if ((dev->class >> 8) != PCI_CLASS_BRIDGE_HOST)
> +		pci_write_config_word(dev, PCI_COMMAND, orig_cmd);
> +
> +	return sz;
> +}
> +
>  static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int
> rom) {
>  	unsigned int pos, reg, next;
> @@ -196,16 +248,13 @@ static void pci_read_bases(struct pci_de
>  		res->name = pci_name(dev);
>  		reg = PCI_BASE_ADDRESS_0 + (pos << 2);
>  		pci_read_config_dword(dev, reg, &l);
> -		pci_write_config_dword(dev, reg, ~0);
> -		pci_read_config_dword(dev, reg, &sz);
> -		pci_write_config_dword(dev, reg, l);
> +		sz = pci_bar_size(dev, reg);
>  		if (!sz || sz == 0xffffffff)
>  			continue;
>  		if (l == 0xffffffff)
>  			l = 0;
>  		raw_sz = sz;
> -		if ((l & PCI_BASE_ADDRESS_SPACE) ==
> -				PCI_BASE_ADDRESS_SPACE_MEMORY) {
> +		if (BAR_IS_MEMORY(l)) {
>  			sz = pci_size(l, sz, (u32)PCI_BASE_ADDRESS_MEM_MASK);
>  			/*
>  			 * For 64bit prefetchable memory sz could be 0, if the
> @@ -229,9 +278,7 @@ static void pci_read_bases(struct pci_de
>  			u32 szhi, lhi;
>
>  			pci_read_config_dword(dev, reg+4, &lhi);
> -			pci_write_config_dword(dev, reg+4, ~0);
> -			pci_read_config_dword(dev, reg+4, &szhi);
> -			pci_write_config_dword(dev, reg+4, lhi);
> +			szhi = pci_bar_size(dev, reg+4);
>  			sz64 = ((u64)szhi << 32) | raw_sz;
>  			l64 = ((u64)lhi << 32) | l;
>  			sz64 = pci_size64(l64, sz64, PCI_BASE_ADDRESS_MEM_MASK);

Looks good to me, though like I said, this probing code could be redone to be 
a little clearer, but that could be done in a separate patch.

Signed-off-by:  Jesse Barnes <jbarnes@xxxxxxxxx>
-
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/