Re: [PATCH 2/3] PCI: endpoint: improve pci_epf_alloc_space()

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Jan 30, 2024 at 08:32:10PM +0100, Niklas Cassel wrote:
> pci_epf_alloc_space() already performs checks on the requested BAR size,
> and will allocate and set epf_bar->size to a size higher than the
> requested BAR size if some constraint deems it necessary.
> 
> However, other than pci_epf_alloc_space() performing these roundups,
> there are checks and roundups in two different places in pci-epf-test.c.
> 
> And further checks are proposed to other endpoint function drivers, see:
> https://lore.kernel.org/linux-pci/20240108151015.2030469-1-Frank.Li@xxxxxxx/
> 
> Having these checks spread out at different places in the EPF driver
> (and potentially in multiple EPF drivers) is not maintainable and makes
> the code hard to follow.
> 
> Since pci_epf_alloc_space() already performs roundups, move the checks and
> roundups performed by pci-epf-test.c to pci_epf_alloc_space().
> 
> Signed-off-by: Niklas Cassel <cassel@xxxxxxxxxx>
> ---
>  drivers/pci/endpoint/functions/pci-epf-test.c |  8 --------
>  drivers/pci/endpoint/pci-epf-core.c           | 10 +++++++++-
>  2 files changed, 9 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/pci/endpoint/functions/pci-epf-test.c b/drivers/pci/endpoint/functions/pci-epf-test.c
> index 15bfa7d83489..981894e40681 100644
> --- a/drivers/pci/endpoint/functions/pci-epf-test.c
> +++ b/drivers/pci/endpoint/functions/pci-epf-test.c
> @@ -841,12 +841,6 @@ static int pci_epf_test_alloc_space(struct pci_epf *epf)
>  	}
>  	test_reg_size = test_reg_bar_size + msix_table_size + pba_size;
>  
> -	if (epc_features->bar_fixed_size[test_reg_bar]) {
> -		if (test_reg_size > bar_size[test_reg_bar])
> -			return -ENOMEM;
> -		test_reg_size = bar_size[test_reg_bar];
> -	}
> -
>  	base = pci_epf_alloc_space(epf, test_reg_size, test_reg_bar,
>  				   epc_features, PRIMARY_INTERFACE);
>  	if (!base) {
> @@ -888,8 +882,6 @@ static void pci_epf_configure_bar(struct pci_epf *epf,
>  		bar_fixed_64bit = !!(epc_features->bar_fixed_64bit & (1 << i));
>  		if (bar_fixed_64bit)
>  			epf_bar->flags |= PCI_BASE_ADDRESS_MEM_TYPE_64;
> -		if (epc_features->bar_fixed_size[i])
> -			bar_size[i] = epc_features->bar_fixed_size[i];
>  	}
>  }
>  
> diff --git a/drivers/pci/endpoint/pci-epf-core.c b/drivers/pci/endpoint/pci-epf-core.c
> index e44f4078fe8b..37d9651d2026 100644
> --- a/drivers/pci/endpoint/pci-epf-core.c
> +++ b/drivers/pci/endpoint/pci-epf-core.c
> @@ -260,6 +260,7 @@ void *pci_epf_alloc_space(struct pci_epf *epf, size_t size, enum pci_barno bar,
>  			  const struct pci_epc_features *epc_features,
>  			  enum pci_epc_interface_type type)
>  {
> +	u64 bar_fixed_size = epc_features->bar_fixed_size[bar];
>  	size_t align = epc_features->align;
>  	struct pci_epf_bar *epf_bar;
>  	dma_addr_t phys_addr;
> @@ -270,7 +271,14 @@ void *pci_epf_alloc_space(struct pci_epf *epf, size_t size, enum pci_barno bar,
>  	if (size < 128)
>  		size = 128;
>  
> -	if (align)
> +	if (bar_fixed_size && size > bar_fixed_size) {
> +		dev_err(dev, "requested BAR size is larger than fixed size\n");
> +		return NULL;
> +	}
> +
> +	if (bar_fixed_size)
> +		size = bar_fixed_size;
> +	else if (align)
>  		size = ALIGN(size, align);
>  	else
>  		size = roundup_pow_of_two(size);

We actually need to perform the alignment even for fixed size BARs too,
since some platforms have fixed_size_bars that are smaller than the
iATU MIN REGION.

Will fix in v2.


Kind regards,
Niklas




[Index of Archives]     [DMA Engine]     [Linux Coverity]     [Linux USB]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Greybus]

  Powered by Linux