PCI: dwc: Fix setting error return on MSI DMA mapping failure

[PATCH v2] PCI: dwc: Fix setting error return on MSI DMA mapping failure

Posted by Jiantao Zhang 4 years, 3 months ago

When dma_mapping_error() returns error because of no enough memory,
but dw_pcie_host_init() returns success, which will mislead the callers.

Fixes: 07940c369a6b ("PCI: dwc: Fix MSI page leakage in suspend/resume")
Signed-off-by: Jianrong Zhang <zhangjianrong5@huawei.com>
Signed-off-by: Jiantao Zhang <water.zhangjiantao@huawei.com>
---
   drivers/pci/controller/dwc/pcie-designware-host.c | 3 ++-
   1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c b/drivers/pci/controller/dwc/pcie-designware-host.c
index f4755f3a03be..9dcb51728dd1 100644
--- a/drivers/pci/controller/dwc/pcie-designware-host.c
+++ b/drivers/pci/controller/dwc/pcie-designware-host.c
@@ -390,7 +390,8 @@ int dw_pcie_host_init(struct pcie_port *pp)
                                 sizeof(pp->msi_msg),
                                 DMA_FROM_DEVICE,
                                 DMA_ATTR_SKIP_CPU_SYNC);
-            if (dma_mapping_error(pci->dev, pp->msi_data)) {
+            ret = dma_mapping_error(pci->dev, pp->msi_data);
+            if (ret) {
                   dev_err(pci->dev, "Failed to map MSI data\n");
                   pp->msi_data = 0;
                   goto err_free_msi;


--
2.17.1

Jiantao Zhang


On Tue, Mar 8, 2022 at 9:45 AM Zhangjiantao(Kirin,Nanjing)
<water.zhangjiantao@huawei.com> wrote:
> When dma_mapping_error() returns error because of no enough memory available, dw_pcie_host_init() will return success, which will mislead the callers.

Again, wrap long lines. And please reword the subject as I previously 
suggested.

It should be 'PATCH v2', not 'PATCH2' in the subject. 'git send-email
-v2' does that for you.

> Fixes: 07940c369a6b ("PCI: dwc: Fix MSI page leakage in suspend/resume")
>

Drop the blank line here.


> Signed-off-by: Jiantao Zhang <water.zhangjiantao@huawei.com>

As you are sending the patch, your S-o-b goes last. And the author
name still doesn't match here: 'Jiantao Zhang' vs.
'Zhangjiantao(Kirin,Nanjing)'

> Signed-off-by: Jianrong Zhang <zhangjianrong5@huawei.com>
> ---
>    drivers/pci/controller/dwc/pcie-designware-host.c | 3 ++-
>    1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c
> b/drivers/pci/controller/dwc/pcie-designware-host.c
> index f4755f3a03be..9dcb51728dd1 100644
> --- a/drivers/pci/controller/dwc/pcie-designware-host.c
> +++ b/drivers/pci/controller/dwc/pcie-designware-host.c
> @@ -390,7 +390,8 @@ int dw_pcie_host_init(struct pcie_port *pp)
>                                  sizeof(pp->msi_msg),
>                                  DMA_FROM_DEVICE,
>                                  DMA_ATTR_SKIP_CPU_SYNC);
> -            if (dma_mapping_error(pci->dev, pp->msi_data)) {
> +            ret = dma_mapping_error(pci->dev, pp->msi_data);
> +            if (ret) {
>                    dev_err(pci->dev, "Failed to map MSI data\n");
>                    pp->msi_data = 0;
>                    goto err_free_msi;
> --
> 2.17.1
>
> Jiantao Zhang
>
> On 2022/3/5 5:39, Bjorn Helgaas wrote:
>> On Fri, Mar 04, 2022 at 12:06:10PM +0000, Zhangjiantao(Kirin,Nanjing) wrote:

All this should not be in new versions of patches.

>>> When dma_mapping_error returns error because of no enough memory available, dw_pcie_host_init will return success, which will mislead the callers.
>> Wrap to fit in 75 columns, add "()" after function names.
>>> Signed-off-by: zhangjianrong <zhangjianrong5@huawei.com>
>>> Signed-off-by: zhangjiantao <water.zhangjiantao@huawei.com>
>>>
>>> ---
>>>     drivers/pci/controller/dwc/pcie-designware-host.c | 1 +
>>>     1 file changed, 1 insertion(+)
>>>
>>> diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c
>>> b/drivers/pci/controller/dwc/pcie-designware-host.c
>>> index f4755f3a03be..ac691d733848 100644
>>> --- a/drivers/pci/controller/dwc/pcie-designware-host.c
>>> +++ b/drivers/pci/controller/dwc/pcie-designware-host.c
>>> @@ -393,6 +393,7 @@ int dw_pcie_host_init(struct pcie_port *pp)
>>>                 if (dma_mapping_error(pci->dev, pp->msi_data)) {
>>>                     dev_err(pci->dev, "Failed to map MSI data\n");
>>>                     pp->msi_data = 0;
>>> +                ret = -1;
>> We should save the return value from dma_mapping_error() and return
>> *that*, not -1.
>>
>> Whitespace error (indent with tabs, same as the surrounding code).
>>
>>>                     goto err_free_msi;
>>>                 }
>>>             }
>>> --
>>> 2.17.1
>>>
>> .

Re: [PATCH v2] PCI: dwc: Fix setting error return on MSI DMA mapping failure

Posted by Rob Herring 4 years, 3 months ago

On Wed, Mar 09, 2022 at 08:01:04PM +0800, Jiantao Zhang wrote:
> When dma_mapping_error() returns error because of no enough memory,
> but dw_pcie_host_init() returns success, which will mislead the callers.
> 
> Fixes: 07940c369a6b ("PCI: dwc: Fix MSI page leakage in suspend/resume")
> Signed-off-by: Jianrong Zhang <zhangjianrong5@huawei.com>
> Signed-off-by: Jiantao Zhang <water.zhangjiantao@huawei.com>
> ---
>   drivers/pci/controller/dwc/pcie-designware-host.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)

Reviewed-by: Rob Herring <robh@kernel.org>

Re: [PATCH v2] PCI: dwc: Fix setting error return on MSI DMA mapping failure

Posted by Lorenzo Pieralisi 4 years, 2 months ago

On Wed, 9 Mar 2022 20:01:04 +0800, Jiantao Zhang wrote:
> When dma_mapping_error() returns error because of no enough memory,
> but dw_pcie_host_init() returns success, which will mislead the callers.
> 
> 

Applied to pci/dwc, thanks!

[1/1] PCI: dwc: Fix setting error return on MSI DMA mapping failure
      https://git.kernel.org/lpieralisi/pci/c/88557685cd

Thanks,
Lorenzo