| Message ID | 20240122115749.67682-2-dmantipov@yandex.ru (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Kalle Valo |
| Headers | show |
| Series | [1/2,v3] wifi: brcmfmac: handle possible completion timeouts | expand |
On January 22, 2024 12:59:03 PM Dmitry Antipov <dmantipov@yandex.ru> wrote: > Switch to newer 'pci_{alloc,feee}_irq_vectors()' API and handle > possible errors in 'brcmf_pcie_request_irq()'. Compile tested only. > > Found by Linux Verification Center (linuxtesting.org) with SVACE. See comment below... Reviewed-by: Arend van Spriel<arend.vanspriel@broadcom.com> > Suggested-by: Bjorn Helgaas <helgaas@kernel.org> > Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru> > --- > v3: switch to 'pci_{alloc,feee}_irq_vectors()' per Bjorn's review > v2: rebase against wireless-next tree > --- > .../net/wireless/broadcom/brcm80211/brcmfmac/pcie.c | 10 +++++++--- > 1 file changed, 7 insertions(+), 3 deletions(-) > > diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c > b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c > index 80220685f5e4..0f77d94f34a3 100644 > --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c > +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c > @@ -965,6 +965,7 @@ static irqreturn_t brcmf_pcie_isr_thread(int irq, void > *arg) > > static int brcmf_pcie_request_irq(struct brcmf_pciedev_info *devinfo) > { > + int ret; > struct pci_dev *pdev = devinfo->pdev; > struct brcmf_bus *bus = dev_get_drvdata(&pdev->dev); > > @@ -972,11 +973,14 @@ static int brcmf_pcie_request_irq(struct > brcmf_pciedev_info *devinfo) > > brcmf_dbg(PCIE, "Enter\n"); > > - pci_enable_msi(pdev); > + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); > + if (ret < 0) > + return ret; > + > if (request_threaded_irq(pdev->irq, Reading the kerneldoc for pci_alloc_irq_vectors() you should use the helper function pci_irq_vector() instead of pdev->irq brcmf_pcie_quick_check_isr, > brcmf_pcie_isr_thread, IRQF_SHARED, > "brcmf_pcie_intr", devinfo)) { > - pci_disable_msi(pdev); > + pci_free_irq_vectors(pdev); > brcmf_err(bus, "Failed to request IRQ %d\n", pdev->irq); > return -EIO; > } > @@ -997,7 +1001,7 @@ static void brcmf_pcie_release_irq(struct > brcmf_pciedev_info *devinfo) > > brcmf_pcie_intr_disable(devinfo); > free_irq(pdev->irq, devinfo); > - pci_disable_msi(pdev); > + pci_free_irq_vectors(pdev); > > msleep(50); > count = 0; > -- > 2.43.0
On Mon, Jan 22, 2024 at 02:57:25PM +0300, Dmitry Antipov wrote: > Switch to newer 'pci_{alloc,feee}_irq_vectors()' API and handle > possible errors in 'brcmf_pcie_request_irq()'. Compile tested only. If you have occasion to update this, possibly s/PCIE/PCI/ in the subject, since this is generic to PCI and PCIe. Bjorn
On 1/22/24 21:20, Bjorn Helgaas wrote: > If you have occasion to update this, possibly s/PCIE/PCI/ in the > subject, since this is generic to PCI and PCIe. OK. BTW if we're touching this anyway, shouldn't we hook into the generic device framework and switch to devm_request_threaded_irq()/devm_free_irq() as well? Dmitry
On January 24, 2024 7:22:58 AM Dmitry Antipov <dmantipov@yandex.ru> wrote: > On 1/22/24 21:20, Bjorn Helgaas wrote: > >> If you have occasion to update this, possibly s/PCIE/PCI/ in the >> subject, since this is generic to PCI and PCIe. > > OK. BTW if we're touching this anyway, shouldn't we hook into the generic > device > framework and switch to devm_request_threaded_irq()/devm_free_irq() as well? Hi Dmitry, Those are not generic device functions. They create device managed resource so devm_free_irq() is implicitly invoked upon device detach. So replacing free_irq() with devm_free_irq() is not very meaningful. There are some devm_*() usages in the driver, but it's not used commonly. However, feel free to add it. Regards, Arend > Dmitry
diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c index 80220685f5e4..0f77d94f34a3 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c @@ -965,6 +965,7 @@ static irqreturn_t brcmf_pcie_isr_thread(int irq, void *arg) static int brcmf_pcie_request_irq(struct brcmf_pciedev_info *devinfo) { + int ret; struct pci_dev *pdev = devinfo->pdev; struct brcmf_bus *bus = dev_get_drvdata(&pdev->dev); @@ -972,11 +973,14 @@ static int brcmf_pcie_request_irq(struct brcmf_pciedev_info *devinfo) brcmf_dbg(PCIE, "Enter\n"); - pci_enable_msi(pdev); + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) + return ret; + if (request_threaded_irq(pdev->irq, brcmf_pcie_quick_check_isr, brcmf_pcie_isr_thread, IRQF_SHARED, "brcmf_pcie_intr", devinfo)) { - pci_disable_msi(pdev); + pci_free_irq_vectors(pdev); brcmf_err(bus, "Failed to request IRQ %d\n", pdev->irq); return -EIO; } @@ -997,7 +1001,7 @@ static void brcmf_pcie_release_irq(struct brcmf_pciedev_info *devinfo) brcmf_pcie_intr_disable(devinfo); free_irq(pdev->irq, devinfo); - pci_disable_msi(pdev); + pci_free_irq_vectors(pdev); msleep(50); count = 0;
Switch to newer 'pci_{alloc,feee}_irq_vectors()' API and handle possible errors in 'brcmf_pcie_request_irq()'. Compile tested only. Found by Linux Verification Center (linuxtesting.org) with SVACE. Suggested-by: Bjorn Helgaas <helgaas@kernel.org> Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru> --- v3: switch to 'pci_{alloc,feee}_irq_vectors()' per Bjorn's review v2: rebase against wireless-next tree --- .../net/wireless/broadcom/brcm80211/brcmfmac/pcie.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-)