| Message ID | 20240626-vfio_pci_mmap-v4-2-7f038870f022@linux.ibm.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | vfio/pci: s390: Fix issues preventing VFIO_PCI_MMAP=y for s390 and enable it | expand |
On Wed, 2024-06-26 at 13:15 +0200, Niklas Schnelle wrote: > On s390 there is a virtual PCI device called ISM which has a few > pecularities. For one it claims to have a 256 TiB PCI BAR which leads to > any attempt to ioremap() in its entirety failing. This is problematic > since mapping the whole BAR is the default behavior of for example QEMU > with VFIO_PCI_MMAP enabled. > > Even if one tried to map this BAR only partially the mapping would not > be usable on systems with MIO support enabled unless using the function > handle based PCI instructions directly. This is because of another > oddity in that this virtual PCI device does not support the newer memory > I/O (MIO) PCI instructions and legacy PCI instructions are not > accessible through writeq()/readq() or by user-space when MIO is in use. > > Indicate that ISM's BAR is special and does not conform to PCI BAR > expectations by setting pdev->non_compliant_bars such that drivers not > specifically developed for ISM can easily ignore it. To this end add > basic PCI quirks support modeled after x86's arch/x86/pci/fixup.c and > move the ISM device's PCI ID to the common header to make it accessible. > Also enable CONFIG_PCI_QUIRKS whenever CONFIG_PCI is enabled. > > Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com> > --- > Fixed the '-' instead of '->' typo in the patch subject locally. Chances are we get a v5 anyway. Thanks, Niklas
On 26.06.24 13:20, Niklas Schnelle wrote: > On Wed, 2024-06-26 at 13:15 +0200, Niklas Schnelle wrote: >> On s390 there is a virtual PCI device called ISM which has a few >> pecularities. For one it claims to have a 256 TiB PCI BAR which leads to >> any attempt to ioremap() in its entirety failing. This is problematic >> since mapping the whole BAR is the default behavior of for example QEMU >> with VFIO_PCI_MMAP enabled. >> >> Even if one tried to map this BAR only partially the mapping would not >> be usable on systems with MIO support enabled unless using the function >> handle based PCI instructions directly. This is because of another >> oddity in that this virtual PCI device does not support the newer memory >> I/O (MIO) PCI instructions and legacy PCI instructions are not >> accessible through writeq()/readq() or by user-space when MIO is in use. >> >> Indicate that ISM's BAR is special and does not conform to PCI BAR >> expectations by setting pdev->non_compliant_bars such that drivers not >> specifically developed for ISM can easily ignore it. To this end add >> basic PCI quirks support modeled after x86's arch/x86/pci/fixup.c and >> move the ISM device's PCI ID to the common header to make it accessible. >> Also enable CONFIG_PCI_QUIRKS whenever CONFIG_PCI is enabled. >> >> Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com> >> --- >> > > > Fixed the '-' instead of '->' typo in the patch subject locally. > Chances are we get a v5 anyway. > > Thanks, > Niklas > For a v5: I am not an expert in English grammar, but I think some commas (',') may help with reading your very good commit messages. Acked-by: Alexandra Winter <wintera@linux.ibm.com>
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index c59d2b54df49..8332ba71465e 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -41,9 +41,6 @@ config AUDIT_ARCH config NO_IOPORT_MAP def_bool y -config PCI_QUIRKS - def_bool n - config ARCH_SUPPORTS_UPROBES def_bool y @@ -234,6 +231,7 @@ config S390 select PCI_DOMAINS if PCI select PCI_MSI if PCI select PCI_MSI_ARCH_FALLBACKS if PCI_MSI + select PCI_QUIRKS if PCI select SPARSE_IRQ select SWIOTLB select SYSCTL_EXCEPTION_TRACE diff --git a/arch/s390/pci/Makefile b/arch/s390/pci/Makefile index 0547a10406e7..03d96c76c21a 100644 --- a/arch/s390/pci/Makefile +++ b/arch/s390/pci/Makefile @@ -5,5 +5,5 @@ obj-$(CONFIG_PCI) += pci.o pci_irq.o pci_clp.o pci_sysfs.o \ pci_event.o pci_debug.o pci_insn.o pci_mmio.o \ - pci_bus.o pci_kvm_hook.o + pci_bus.o pci_kvm_hook.o pci_fixup.o obj-$(CONFIG_PCI_IOV) += pci_iov.o diff --git a/arch/s390/pci/pci_fixup.c b/arch/s390/pci/pci_fixup.c new file mode 100644 index 000000000000..61045348f20a --- /dev/null +++ b/arch/s390/pci/pci_fixup.c @@ -0,0 +1,23 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Exceptions for specific devices, + * + * Copyright IBM Corp. 2024 + * + * Author(s): + * Niklas Schnelle <schnelle@linux.ibm.com> + */ +#include <linux/pci.h> + +static void zpci_ism_bar_no_mmap(struct pci_dev *pdev) +{ + /* + * ISM's BAR is special. Drivers written for ISM know + * how to handle this but others need to be aware of their + * special nature e.g. to prevent attempts to mmap() it. + */ + pdev->non_compliant_bars = 1; +} +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_IBM, + PCI_DEVICE_ID_IBM_ISM, + zpci_ism_bar_no_mmap); diff --git a/drivers/s390/net/ism_drv.c b/drivers/s390/net/ism_drv.c index e36e3ea165d3..d32633ed9fa8 100644 --- a/drivers/s390/net/ism_drv.c +++ b/drivers/s390/net/ism_drv.c @@ -20,7 +20,6 @@ MODULE_DESCRIPTION("ISM driver for s390"); MODULE_LICENSE("GPL"); -#define PCI_DEVICE_ID_IBM_ISM 0x04ED #define DRV_NAME "ism" static const struct pci_device_id ism_device_table[] = { diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 942a587bb97e..d730fce2113f 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -517,6 +517,7 @@ #define PCI_DEVICE_ID_IBM_ICOM_V2_ONE_PORT_RVX_ONE_PORT_MDM 0x0251 #define PCI_DEVICE_ID_IBM_ICOM_V2_ONE_PORT_RVX_ONE_PORT_MDM_PCIE 0x0361 #define PCI_DEVICE_ID_IBM_ICOM_FOUR_PORT_MODEL 0x252 +#define PCI_DEVICE_ID_IBM_ISM 0x04ED #define PCI_SUBVENDOR_ID_IBM 0x1014 #define PCI_SUBDEVICE_ID_IBM_SATURN_SERIAL_ONE_PORT 0x03d4
On s390 there is a virtual PCI device called ISM which has a few pecularities. For one it claims to have a 256 TiB PCI BAR which leads to any attempt to ioremap() in its entirety failing. This is problematic since mapping the whole BAR is the default behavior of for example QEMU with VFIO_PCI_MMAP enabled. Even if one tried to map this BAR only partially the mapping would not be usable on systems with MIO support enabled unless using the function handle based PCI instructions directly. This is because of another oddity in that this virtual PCI device does not support the newer memory I/O (MIO) PCI instructions and legacy PCI instructions are not accessible through writeq()/readq() or by user-space when MIO is in use. Indicate that ISM's BAR is special and does not conform to PCI BAR expectations by setting pdev->non_compliant_bars such that drivers not specifically developed for ISM can easily ignore it. To this end add basic PCI quirks support modeled after x86's arch/x86/pci/fixup.c and move the ISM device's PCI ID to the common header to make it accessible. Also enable CONFIG_PCI_QUIRKS whenever CONFIG_PCI is enabled. Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com> --- arch/s390/Kconfig | 4 +--- arch/s390/pci/Makefile | 2 +- arch/s390/pci/pci_fixup.c | 23 +++++++++++++++++++++++ drivers/s390/net/ism_drv.c | 1 - include/linux/pci_ids.h | 1 + 5 files changed, 26 insertions(+), 5 deletions(-)