| Message ID | 20200508130646.23939-5-kishon@ti.com (mailing list archive) |
|---|---|
| State | Changes Requested, archived |
| Delegated to: | Lorenzo Pieralisi |
| Headers | show |
| Series | PCI: cadence: Deprecate inbound/outbound specific bindings | expand |
Hi RobH, Robin, On 5/8/2020 6:36 PM, Kishon Vijay Abraham I wrote: > Cadence PCIe core driver (host mode) uses "cdns,no-bar-match-nbits" > property to configure the number of bits passed through from PCIe > address to internal address in Inbound Address Translation register. > This only used the NO MATCH BAR. > > However standard PCI dt-binding already defines "dma-ranges" to > describe the address ranges accessible by PCIe controller. Add support > in Cadence PCIe host driver to parse dma-ranges and configure the > inbound regions for BAR0, BAR1 and NO MATCH BAR. Cadence IP specifies > maximum size for BAR0 as 256GB, maximum size for BAR1 as 2 GB, so if > the dma-ranges specifies a size larger than the maximum allowed, the > driver will split and configure the BARs. > > Legacy device tree binding compatibility is maintained by retaining > support for "cdns,no-bar-match-nbits". Do you have any comments for this patch? Tom, where you able to test this in Cadence platform? Thanks Kishon > > Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> > --- > .../controller/cadence/pcie-cadence-host.c | 141 ++++++++++++++++-- > drivers/pci/controller/cadence/pcie-cadence.h | 17 ++- > 2 files changed, 141 insertions(+), 17 deletions(-) > > diff --git a/drivers/pci/controller/cadence/pcie-cadence-host.c b/drivers/pci/controller/cadence/pcie-cadence-host.c > index 6ecebb79057a..2485ecd8434d 100644 > --- a/drivers/pci/controller/cadence/pcie-cadence-host.c > +++ b/drivers/pci/controller/cadence/pcie-cadence-host.c > @@ -11,6 +11,12 @@ > > #include "pcie-cadence.h" > > +static u64 cdns_rp_bar_max_size[] = { > + [RP_BAR0] = _ULL(128 * SZ_2G), > + [RP_BAR1] = SZ_2G, > + [RP_NO_BAR] = SZ_64T, > +}; > + > void __iomem *cdns_pci_map_bus(struct pci_bus *bus, unsigned int devfn, > int where) > { > @@ -106,6 +112,117 @@ static int cdns_pcie_host_init_root_port(struct cdns_pcie_rc *rc) > return 0; > } > > +static void cdns_pcie_host_bar_ib_config(struct cdns_pcie_rc *rc, > + enum cdns_pcie_rp_bar bar, > + u64 cpu_addr, u32 aperture) > +{ > + struct cdns_pcie *pcie = &rc->pcie; > + u32 addr0, addr1; > + > + addr0 = CDNS_PCIE_AT_IB_RP_BAR_ADDR0_NBITS(aperture) | > + (lower_32_bits(cpu_addr) & GENMASK(31, 8)); > + addr1 = upper_32_bits(cpu_addr); > + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR0(bar), addr0); > + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR1(bar), addr1); > +} > + > +static int cdns_pcie_host_bar_config(struct cdns_pcie_rc *rc, > + struct resource_entry *entry, > + enum cdns_pcie_rp_bar *index) > +{ > + u64 cpu_addr, pci_addr, size, winsize; > + struct cdns_pcie *pcie = &rc->pcie; > + struct device *dev = pcie->dev; > + enum cdns_pcie_rp_bar bar; > + unsigned long flags; > + u32 aperture; > + u32 value; > + > + cpu_addr = entry->res->start; > + flags = entry->res->flags; > + pci_addr = entry->res->start - entry->offset; > + size = resource_size(entry->res); > + bar = *index; > + > + if (entry->offset) { > + dev_err(dev, "Cannot map PCI addr: %llx to CPU addr: %llx\n", > + pci_addr, cpu_addr); > + return -EINVAL; > + } > + > + value = cdns_pcie_readl(pcie, CDNS_PCIE_LM_RC_BAR_CFG); > + while (size > 0) { > + if (bar > RP_NO_BAR) { > + dev_err(dev, "Failed to map inbound regions!\n"); > + return -EINVAL; > + } > + > + winsize = size; > + if (size > cdns_rp_bar_max_size[bar]) > + winsize = cdns_rp_bar_max_size[bar]; > + > + aperture = ilog2(winsize); > + > + cdns_pcie_host_bar_ib_config(rc, bar, cpu_addr, aperture); > + > + if (bar == RP_NO_BAR) > + break; > + > + if (winsize + cpu_addr >= SZ_4G) { > + if (!(flags & IORESOURCE_PREFETCH)) > + value |= LM_RC_BAR_CFG_CTRL_MEM_64BITS(bar); > + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_64BITS(bar); > + } else { > + if (!(flags & IORESOURCE_PREFETCH)) > + value |= LM_RC_BAR_CFG_CTRL_MEM_32BITS(bar); > + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_32BITS(bar); > + } > + > + value |= LM_RC_BAR_CFG_APERTURE(bar, aperture); > + > + size -= winsize; > + cpu_addr += winsize; > + bar++; > + } > + cdns_pcie_writel(pcie, CDNS_PCIE_LM_RC_BAR_CFG, value); > + *index = bar; > + > + return 0; > +} > + > +static int cdns_pcie_host_map_dma_ranges(struct cdns_pcie_rc *rc) > +{ > + enum cdns_pcie_rp_bar bar = RP_BAR0; > + struct cdns_pcie *pcie = &rc->pcie; > + struct device *dev = pcie->dev; > + struct device_node *np = dev->of_node; > + struct pci_host_bridge *bridge; > + struct resource_entry *entry; > + u32 no_bar_nbits = 32; > + int err; > + > + bridge = pci_host_bridge_from_priv(rc); > + if (!bridge) > + return -ENOMEM; > + > + if (list_empty(&bridge->dma_ranges)) { > + of_property_read_u32(np, "cdns,no-bar-match-nbits", > + &no_bar_nbits); > + cdns_pcie_host_bar_ib_config(rc, RP_NO_BAR, 0x0, no_bar_nbits); > + return 0; > + } > + > + resource_list_for_each_entry(entry, &bridge->dma_ranges) { > + err = cdns_pcie_host_bar_config(rc, entry, &bar); > + if (err) { > + dev_err(dev, "Fail to configure IB using dma-ranges\n"); > + return err; > + } > + } > + > + return 0; > +} > + > static int cdns_pcie_host_init_address_translation(struct cdns_pcie_rc *rc) > { > struct cdns_pcie *pcie = &rc->pcie; > @@ -160,16 +277,9 @@ static int cdns_pcie_host_init_address_translation(struct cdns_pcie_rc *rc) > r++; > } > > - /* > - * Set Root Port no BAR match Inbound Translation registers: > - * needed for MSI and DMA. > - * Root Port BAR0 and BAR1 are disabled, hence no need to set their > - * inbound translation registers. > - */ > - addr0 = CDNS_PCIE_AT_IB_RP_BAR_ADDR0_NBITS(rc->no_bar_nbits); > - addr1 = 0; > - cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR0(RP_NO_BAR), addr0); > - cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR1(RP_NO_BAR), addr1); > + err = cdns_pcie_host_map_dma_ranges(rc); > + if (err) > + return err; > > return 0; > } > @@ -179,10 +289,16 @@ static int cdns_pcie_host_init(struct device *dev, > struct cdns_pcie_rc *rc) > { > struct resource *bus_range = NULL; > + struct pci_host_bridge *bridge; > int err; > > + bridge = pci_host_bridge_from_priv(rc); > + if (!bridge) > + return -ENOMEM; > + > /* Parse our PCI ranges and request their resources */ > - err = pci_parse_request_of_pci_ranges(dev, resources, NULL, &bus_range); > + err = pci_parse_request_of_pci_ranges(dev, resources, > + &bridge->dma_ranges, &bus_range); > if (err) > return err; > > @@ -239,9 +355,6 @@ int cdns_pcie_host_setup(struct cdns_pcie_rc *rc) > pcie = &rc->pcie; > pcie->is_rc = true; > > - rc->no_bar_nbits = 32; > - of_property_read_u32(np, "cdns,no-bar-match-nbits", &rc->no_bar_nbits); > - > rc->vendor_id = 0xffff; > of_property_read_u32(np, "vendor-id", &rc->vendor_id); > > diff --git a/drivers/pci/controller/cadence/pcie-cadence.h b/drivers/pci/controller/cadence/pcie-cadence.h > index f349f5828a58..336237a7025c 100644 > --- a/drivers/pci/controller/cadence/pcie-cadence.h > +++ b/drivers/pci/controller/cadence/pcie-cadence.h > @@ -92,6 +92,20 @@ > #define CDNS_PCIE_LM_BAR_CFG_CTRL_MEM_64BITS 0x6 > #define CDNS_PCIE_LM_BAR_CFG_CTRL_PREFETCH_MEM_64BITS 0x7 > > +#define LM_RC_BAR_CFG_CTRL_DISABLED(bar) \ > + (CDNS_PCIE_LM_BAR_CFG_CTRL_DISABLED << (((bar) * 8) + 6)) > +#define LM_RC_BAR_CFG_CTRL_IO_32BITS(bar) \ > + (CDNS_PCIE_LM_BAR_CFG_CTRL_IO_32BITS << (((bar) * 8) + 6)) > +#define LM_RC_BAR_CFG_CTRL_MEM_32BITS(bar) \ > + (CDNS_PCIE_LM_BAR_CFG_CTRL_MEM_32BITS << (((bar) * 8) + 6)) > +#define LM_RC_BAR_CFG_CTRL_PREF_MEM_32BITS(bar) \ > + (CDNS_PCIE_LM_BAR_CFG_CTRL_PREFETCH_MEM_32BITS << (((bar) * 8) + 6)) > +#define LM_RC_BAR_CFG_CTRL_MEM_64BITS(bar) \ > + (CDNS_PCIE_LM_BAR_CFG_CTRL_MEM_64BITS << (((bar) * 8) + 6)) > +#define LM_RC_BAR_CFG_CTRL_PREF_MEM_64BITS(bar) \ > + (CDNS_PCIE_LM_BAR_CFG_CTRL_PREFETCH_MEM_64BITS << (((bar) * 8) + 6)) > +#define LM_RC_BAR_CFG_APERTURE(bar, aperture) \ > + (((aperture) - 2) << ((bar) * 8)) > > /* > * Endpoint Function Registers (PCI configuration space for endpoint functions) > @@ -266,8 +280,6 @@ struct cdns_pcie { > * @bus_range: first/last buses behind the PCIe host controller > * @cfg_base: IO mapped window to access the PCI configuration space of a > * single function at a time > - * @no_bar_nbits: Number of bits to keep for inbound (PCIe -> CPU) address > - * translation (nbits sets into the "no BAR match" register) > * @vendor_id: PCI vendor ID > * @device_id: PCI device ID > */ > @@ -276,7 +288,6 @@ struct cdns_pcie_rc { > struct resource *cfg_res; > struct resource *bus_range; > void __iomem *cfg_base; > - u32 no_bar_nbits; > u32 vendor_id; > u32 device_id; > }; >
On Fri, May 8, 2020 at 7:07 AM Kishon Vijay Abraham I <kishon@ti.com> wrote: > > Cadence PCIe core driver (host mode) uses "cdns,no-bar-match-nbits" > property to configure the number of bits passed through from PCIe > address to internal address in Inbound Address Translation register. > This only used the NO MATCH BAR. > > However standard PCI dt-binding already defines "dma-ranges" to > describe the address ranges accessible by PCIe controller. Add support > in Cadence PCIe host driver to parse dma-ranges and configure the > inbound regions for BAR0, BAR1 and NO MATCH BAR. Cadence IP specifies > maximum size for BAR0 as 256GB, maximum size for BAR1 as 2 GB, so if > the dma-ranges specifies a size larger than the maximum allowed, the > driver will split and configure the BARs. Would be useful to know what your dma-ranges contains now. > Legacy device tree binding compatibility is maintained by retaining > support for "cdns,no-bar-match-nbits". > > Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> > --- > .../controller/cadence/pcie-cadence-host.c | 141 ++++++++++++++++-- > drivers/pci/controller/cadence/pcie-cadence.h | 17 ++- > 2 files changed, 141 insertions(+), 17 deletions(-) > > diff --git a/drivers/pci/controller/cadence/pcie-cadence-host.c b/drivers/pci/controller/cadence/pcie-cadence-host.c > index 6ecebb79057a..2485ecd8434d 100644 > --- a/drivers/pci/controller/cadence/pcie-cadence-host.c > +++ b/drivers/pci/controller/cadence/pcie-cadence-host.c > @@ -11,6 +11,12 @@ > > #include "pcie-cadence.h" > > +static u64 cdns_rp_bar_max_size[] = { > + [RP_BAR0] = _ULL(128 * SZ_2G), > + [RP_BAR1] = SZ_2G, > + [RP_NO_BAR] = SZ_64T, > +}; > + > void __iomem *cdns_pci_map_bus(struct pci_bus *bus, unsigned int devfn, > int where) > { > @@ -106,6 +112,117 @@ static int cdns_pcie_host_init_root_port(struct cdns_pcie_rc *rc) > return 0; > } > > +static void cdns_pcie_host_bar_ib_config(struct cdns_pcie_rc *rc, > + enum cdns_pcie_rp_bar bar, > + u64 cpu_addr, u32 aperture) > +{ > + struct cdns_pcie *pcie = &rc->pcie; > + u32 addr0, addr1; > + > + addr0 = CDNS_PCIE_AT_IB_RP_BAR_ADDR0_NBITS(aperture) | > + (lower_32_bits(cpu_addr) & GENMASK(31, 8)); > + addr1 = upper_32_bits(cpu_addr); > + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR0(bar), addr0); > + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR1(bar), addr1); > +} > + > +static int cdns_pcie_host_bar_config(struct cdns_pcie_rc *rc, > + struct resource_entry *entry, > + enum cdns_pcie_rp_bar *index) > +{ > + u64 cpu_addr, pci_addr, size, winsize; > + struct cdns_pcie *pcie = &rc->pcie; > + struct device *dev = pcie->dev; > + enum cdns_pcie_rp_bar bar; > + unsigned long flags; > + u32 aperture; > + u32 value; > + > + cpu_addr = entry->res->start; > + flags = entry->res->flags; > + pci_addr = entry->res->start - entry->offset; > + size = resource_size(entry->res); > + bar = *index; > + > + if (entry->offset) { > + dev_err(dev, "Cannot map PCI addr: %llx to CPU addr: %llx\n", > + pci_addr, cpu_addr); Would be a bit more clear to say PCI addr must equal CPU addr. > + return -EINVAL; > + } > + > + value = cdns_pcie_readl(pcie, CDNS_PCIE_LM_RC_BAR_CFG); > + while (size > 0) { > + if (bar > RP_NO_BAR) { > + dev_err(dev, "Failed to map inbound regions!\n"); > + return -EINVAL; > + } > + > + winsize = size; > + if (size > cdns_rp_bar_max_size[bar]) > + winsize = cdns_rp_bar_max_size[bar]; > + > + aperture = ilog2(winsize); > + > + cdns_pcie_host_bar_ib_config(rc, bar, cpu_addr, aperture); > + > + if (bar == RP_NO_BAR) > + break; > + > + if (winsize + cpu_addr >= SZ_4G) { > + if (!(flags & IORESOURCE_PREFETCH)) > + value |= LM_RC_BAR_CFG_CTRL_MEM_64BITS(bar); > + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_64BITS(bar); > + } else { > + if (!(flags & IORESOURCE_PREFETCH)) > + value |= LM_RC_BAR_CFG_CTRL_MEM_32BITS(bar); > + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_32BITS(bar); > + } > + > + value |= LM_RC_BAR_CFG_APERTURE(bar, aperture); > + > + size -= winsize; > + cpu_addr += winsize; > + bar++; > + } > + cdns_pcie_writel(pcie, CDNS_PCIE_LM_RC_BAR_CFG, value); > + *index = bar; > + > + return 0; > +} > + > +static int cdns_pcie_host_map_dma_ranges(struct cdns_pcie_rc *rc) > +{ > + enum cdns_pcie_rp_bar bar = RP_BAR0; > + struct cdns_pcie *pcie = &rc->pcie; > + struct device *dev = pcie->dev; > + struct device_node *np = dev->of_node; > + struct pci_host_bridge *bridge; > + struct resource_entry *entry; > + u32 no_bar_nbits = 32; > + int err; > + > + bridge = pci_host_bridge_from_priv(rc); > + if (!bridge) > + return -ENOMEM; > + > + if (list_empty(&bridge->dma_ranges)) { > + of_property_read_u32(np, "cdns,no-bar-match-nbits", > + &no_bar_nbits); > + cdns_pcie_host_bar_ib_config(rc, RP_NO_BAR, 0x0, no_bar_nbits); > + return 0; > + } > + > + resource_list_for_each_entry(entry, &bridge->dma_ranges) { > + err = cdns_pcie_host_bar_config(rc, entry, &bar); Seems like this should have some better logic to pick which BAR to use. Something like find the biggest region and then find the smallest BAR that it fits in. Then get the next biggest... > + if (err) { > + dev_err(dev, "Fail to configure IB using dma-ranges\n"); > + return err; > + } > + } > + > + return 0; > +} > +
Hi Rob, On 5/19/2020 10:41 PM, Rob Herring wrote: > On Fri, May 8, 2020 at 7:07 AM Kishon Vijay Abraham I <kishon@ti.com> wrote: >> >> Cadence PCIe core driver (host mode) uses "cdns,no-bar-match-nbits" >> property to configure the number of bits passed through from PCIe >> address to internal address in Inbound Address Translation register. >> This only used the NO MATCH BAR. >> >> However standard PCI dt-binding already defines "dma-ranges" to >> describe the address ranges accessible by PCIe controller. Add support >> in Cadence PCIe host driver to parse dma-ranges and configure the >> inbound regions for BAR0, BAR1 and NO MATCH BAR. Cadence IP specifies >> maximum size for BAR0 as 256GB, maximum size for BAR1 as 2 GB, so if >> the dma-ranges specifies a size larger than the maximum allowed, the >> driver will split and configure the BARs. > > Would be useful to know what your dma-ranges contains now. > > >> Legacy device tree binding compatibility is maintained by retaining >> support for "cdns,no-bar-match-nbits". >> >> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> >> --- >> .../controller/cadence/pcie-cadence-host.c | 141 ++++++++++++++++-- >> drivers/pci/controller/cadence/pcie-cadence.h | 17 ++- >> 2 files changed, 141 insertions(+), 17 deletions(-) >> >> diff --git a/drivers/pci/controller/cadence/pcie-cadence-host.c b/drivers/pci/controller/cadence/pcie-cadence-host.c >> index 6ecebb79057a..2485ecd8434d 100644 >> --- a/drivers/pci/controller/cadence/pcie-cadence-host.c >> +++ b/drivers/pci/controller/cadence/pcie-cadence-host.c >> @@ -11,6 +11,12 @@ >> >> #include "pcie-cadence.h" >> >> +static u64 cdns_rp_bar_max_size[] = { >> + [RP_BAR0] = _ULL(128 * SZ_2G), >> + [RP_BAR1] = SZ_2G, >> + [RP_NO_BAR] = SZ_64T, >> +}; >> + >> void __iomem *cdns_pci_map_bus(struct pci_bus *bus, unsigned int devfn, >> int where) >> { >> @@ -106,6 +112,117 @@ static int cdns_pcie_host_init_root_port(struct cdns_pcie_rc *rc) >> return 0; >> } >> >> +static void cdns_pcie_host_bar_ib_config(struct cdns_pcie_rc *rc, >> + enum cdns_pcie_rp_bar bar, >> + u64 cpu_addr, u32 aperture) >> +{ >> + struct cdns_pcie *pcie = &rc->pcie; >> + u32 addr0, addr1; >> + >> + addr0 = CDNS_PCIE_AT_IB_RP_BAR_ADDR0_NBITS(aperture) | >> + (lower_32_bits(cpu_addr) & GENMASK(31, 8)); >> + addr1 = upper_32_bits(cpu_addr); >> + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR0(bar), addr0); >> + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR1(bar), addr1); >> +} >> + >> +static int cdns_pcie_host_bar_config(struct cdns_pcie_rc *rc, >> + struct resource_entry *entry, >> + enum cdns_pcie_rp_bar *index) >> +{ >> + u64 cpu_addr, pci_addr, size, winsize; >> + struct cdns_pcie *pcie = &rc->pcie; >> + struct device *dev = pcie->dev; >> + enum cdns_pcie_rp_bar bar; >> + unsigned long flags; >> + u32 aperture; >> + u32 value; >> + >> + cpu_addr = entry->res->start; >> + flags = entry->res->flags; >> + pci_addr = entry->res->start - entry->offset; >> + size = resource_size(entry->res); >> + bar = *index; >> + >> + if (entry->offset) { >> + dev_err(dev, "Cannot map PCI addr: %llx to CPU addr: %llx\n", >> + pci_addr, cpu_addr); > > Would be a bit more clear to say PCI addr must equal CPU addr. > >> + return -EINVAL; >> + } >> + >> + value = cdns_pcie_readl(pcie, CDNS_PCIE_LM_RC_BAR_CFG); >> + while (size > 0) { >> + if (bar > RP_NO_BAR) { >> + dev_err(dev, "Failed to map inbound regions!\n"); >> + return -EINVAL; >> + } >> + >> + winsize = size; >> + if (size > cdns_rp_bar_max_size[bar]) >> + winsize = cdns_rp_bar_max_size[bar]; >> + >> + aperture = ilog2(winsize); >> + >> + cdns_pcie_host_bar_ib_config(rc, bar, cpu_addr, aperture); >> + >> + if (bar == RP_NO_BAR) >> + break; >> + >> + if (winsize + cpu_addr >= SZ_4G) { >> + if (!(flags & IORESOURCE_PREFETCH)) >> + value |= LM_RC_BAR_CFG_CTRL_MEM_64BITS(bar); >> + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_64BITS(bar); >> + } else { >> + if (!(flags & IORESOURCE_PREFETCH)) >> + value |= LM_RC_BAR_CFG_CTRL_MEM_32BITS(bar); >> + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_32BITS(bar); >> + } >> + >> + value |= LM_RC_BAR_CFG_APERTURE(bar, aperture); >> + >> + size -= winsize; >> + cpu_addr += winsize; >> + bar++; >> + } >> + cdns_pcie_writel(pcie, CDNS_PCIE_LM_RC_BAR_CFG, value); >> + *index = bar; >> + >> + return 0; >> +} >> + >> +static int cdns_pcie_host_map_dma_ranges(struct cdns_pcie_rc *rc) >> +{ >> + enum cdns_pcie_rp_bar bar = RP_BAR0; >> + struct cdns_pcie *pcie = &rc->pcie; >> + struct device *dev = pcie->dev; >> + struct device_node *np = dev->of_node; >> + struct pci_host_bridge *bridge; >> + struct resource_entry *entry; >> + u32 no_bar_nbits = 32; >> + int err; >> + >> + bridge = pci_host_bridge_from_priv(rc); >> + if (!bridge) >> + return -ENOMEM; >> + >> + if (list_empty(&bridge->dma_ranges)) { >> + of_property_read_u32(np, "cdns,no-bar-match-nbits", >> + &no_bar_nbits); >> + cdns_pcie_host_bar_ib_config(rc, RP_NO_BAR, 0x0, no_bar_nbits); >> + return 0; >> + } >> + >> + resource_list_for_each_entry(entry, &bridge->dma_ranges) { >> + err = cdns_pcie_host_bar_config(rc, entry, &bar); > > Seems like this should have some better logic to pick which BAR to > use. Something like find the biggest region and then find the smallest > BAR that it fits in. Then get the next biggest... Okay, I'll change this something like for each region, find the smallest BAR that it fits in and if there is no BAR big enough to hold the region, split the region to see if can be fitted using multiple BARs. I don't see the purpose of finding the biggest region first since at all times we'll only use the smallest BAR to fit. Thanks Kishon
Rob, On 5/21/2020 9:00 AM, Kishon Vijay Abraham I wrote: > Hi Rob, > > On 5/19/2020 10:41 PM, Rob Herring wrote: >> On Fri, May 8, 2020 at 7:07 AM Kishon Vijay Abraham I <kishon@ti.com> wrote: >>> >>> Cadence PCIe core driver (host mode) uses "cdns,no-bar-match-nbits" >>> property to configure the number of bits passed through from PCIe >>> address to internal address in Inbound Address Translation register. >>> This only used the NO MATCH BAR. >>> >>> However standard PCI dt-binding already defines "dma-ranges" to >>> describe the address ranges accessible by PCIe controller. Add support >>> in Cadence PCIe host driver to parse dma-ranges and configure the >>> inbound regions for BAR0, BAR1 and NO MATCH BAR. Cadence IP specifies >>> maximum size for BAR0 as 256GB, maximum size for BAR1 as 2 GB, so if >>> the dma-ranges specifies a size larger than the maximum allowed, the >>> driver will split and configure the BARs. >> >> Would be useful to know what your dma-ranges contains now. >> >> >>> Legacy device tree binding compatibility is maintained by retaining >>> support for "cdns,no-bar-match-nbits". >>> >>> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> >>> --- >>> .../controller/cadence/pcie-cadence-host.c | 141 ++++++++++++++++-- >>> drivers/pci/controller/cadence/pcie-cadence.h | 17 ++- >>> 2 files changed, 141 insertions(+), 17 deletions(-) >>> >>> diff --git a/drivers/pci/controller/cadence/pcie-cadence-host.c b/drivers/pci/controller/cadence/pcie-cadence-host.c >>> index 6ecebb79057a..2485ecd8434d 100644 >>> --- a/drivers/pci/controller/cadence/pcie-cadence-host.c >>> +++ b/drivers/pci/controller/cadence/pcie-cadence-host.c >>> @@ -11,6 +11,12 @@ >>> >>> #include "pcie-cadence.h" >>> >>> +static u64 cdns_rp_bar_max_size[] = { >>> + [RP_BAR0] = _ULL(128 * SZ_2G), >>> + [RP_BAR1] = SZ_2G, >>> + [RP_NO_BAR] = SZ_64T, >>> +}; >>> + >>> void __iomem *cdns_pci_map_bus(struct pci_bus *bus, unsigned int devfn, >>> int where) >>> { >>> @@ -106,6 +112,117 @@ static int cdns_pcie_host_init_root_port(struct cdns_pcie_rc *rc) >>> return 0; >>> } >>> >>> +static void cdns_pcie_host_bar_ib_config(struct cdns_pcie_rc *rc, >>> + enum cdns_pcie_rp_bar bar, >>> + u64 cpu_addr, u32 aperture) >>> +{ >>> + struct cdns_pcie *pcie = &rc->pcie; >>> + u32 addr0, addr1; >>> + >>> + addr0 = CDNS_PCIE_AT_IB_RP_BAR_ADDR0_NBITS(aperture) | >>> + (lower_32_bits(cpu_addr) & GENMASK(31, 8)); >>> + addr1 = upper_32_bits(cpu_addr); >>> + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR0(bar), addr0); >>> + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR1(bar), addr1); >>> +} >>> + >>> +static int cdns_pcie_host_bar_config(struct cdns_pcie_rc *rc, >>> + struct resource_entry *entry, >>> + enum cdns_pcie_rp_bar *index) >>> +{ >>> + u64 cpu_addr, pci_addr, size, winsize; >>> + struct cdns_pcie *pcie = &rc->pcie; >>> + struct device *dev = pcie->dev; >>> + enum cdns_pcie_rp_bar bar; >>> + unsigned long flags; >>> + u32 aperture; >>> + u32 value; >>> + >>> + cpu_addr = entry->res->start; >>> + flags = entry->res->flags; >>> + pci_addr = entry->res->start - entry->offset; >>> + size = resource_size(entry->res); >>> + bar = *index; >>> + >>> + if (entry->offset) { >>> + dev_err(dev, "Cannot map PCI addr: %llx to CPU addr: %llx\n", >>> + pci_addr, cpu_addr); >> >> Would be a bit more clear to say PCI addr must equal CPU addr. >> >>> + return -EINVAL; >>> + } >>> + >>> + value = cdns_pcie_readl(pcie, CDNS_PCIE_LM_RC_BAR_CFG); >>> + while (size > 0) { >>> + if (bar > RP_NO_BAR) { >>> + dev_err(dev, "Failed to map inbound regions!\n"); >>> + return -EINVAL; >>> + } >>> + >>> + winsize = size; >>> + if (size > cdns_rp_bar_max_size[bar]) >>> + winsize = cdns_rp_bar_max_size[bar]; >>> + >>> + aperture = ilog2(winsize); >>> + >>> + cdns_pcie_host_bar_ib_config(rc, bar, cpu_addr, aperture); >>> + >>> + if (bar == RP_NO_BAR) >>> + break; >>> + >>> + if (winsize + cpu_addr >= SZ_4G) { >>> + if (!(flags & IORESOURCE_PREFETCH)) >>> + value |= LM_RC_BAR_CFG_CTRL_MEM_64BITS(bar); >>> + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_64BITS(bar); >>> + } else { >>> + if (!(flags & IORESOURCE_PREFETCH)) >>> + value |= LM_RC_BAR_CFG_CTRL_MEM_32BITS(bar); >>> + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_32BITS(bar); >>> + } >>> + >>> + value |= LM_RC_BAR_CFG_APERTURE(bar, aperture); >>> + >>> + size -= winsize; >>> + cpu_addr += winsize; >>> + bar++; >>> + } >>> + cdns_pcie_writel(pcie, CDNS_PCIE_LM_RC_BAR_CFG, value); >>> + *index = bar; >>> + >>> + return 0; >>> +} >>> + >>> +static int cdns_pcie_host_map_dma_ranges(struct cdns_pcie_rc *rc) >>> +{ >>> + enum cdns_pcie_rp_bar bar = RP_BAR0; >>> + struct cdns_pcie *pcie = &rc->pcie; >>> + struct device *dev = pcie->dev; >>> + struct device_node *np = dev->of_node; >>> + struct pci_host_bridge *bridge; >>> + struct resource_entry *entry; >>> + u32 no_bar_nbits = 32; >>> + int err; >>> + >>> + bridge = pci_host_bridge_from_priv(rc); >>> + if (!bridge) >>> + return -ENOMEM; >>> + >>> + if (list_empty(&bridge->dma_ranges)) { >>> + of_property_read_u32(np, "cdns,no-bar-match-nbits", >>> + &no_bar_nbits); >>> + cdns_pcie_host_bar_ib_config(rc, RP_NO_BAR, 0x0, no_bar_nbits); >>> + return 0; >>> + } >>> + >>> + resource_list_for_each_entry(entry, &bridge->dma_ranges) { >>> + err = cdns_pcie_host_bar_config(rc, entry, &bar); >> >> Seems like this should have some better logic to pick which BAR to >> use. Something like find the biggest region and then find the smallest >> BAR that it fits in. Then get the next biggest... > > Okay, I'll change this something like for each region, find the smallest BAR > that it fits in and if there is no BAR big enough to hold the region, split the > region to see if can be fitted using multiple BARs. I don't see the purpose of > finding the biggest region first since at all times we'll only use the smallest > BAR to fit. Nevermind, I realized finding the biggest region is useful. I have sent a patch adding support for that. Thanks Kishon
diff --git a/drivers/pci/controller/cadence/pcie-cadence-host.c b/drivers/pci/controller/cadence/pcie-cadence-host.c index 6ecebb79057a..2485ecd8434d 100644 --- a/drivers/pci/controller/cadence/pcie-cadence-host.c +++ b/drivers/pci/controller/cadence/pcie-cadence-host.c @@ -11,6 +11,12 @@ #include "pcie-cadence.h" +static u64 cdns_rp_bar_max_size[] = { + [RP_BAR0] = _ULL(128 * SZ_2G), + [RP_BAR1] = SZ_2G, + [RP_NO_BAR] = SZ_64T, +}; + void __iomem *cdns_pci_map_bus(struct pci_bus *bus, unsigned int devfn, int where) { @@ -106,6 +112,117 @@ static int cdns_pcie_host_init_root_port(struct cdns_pcie_rc *rc) return 0; } +static void cdns_pcie_host_bar_ib_config(struct cdns_pcie_rc *rc, + enum cdns_pcie_rp_bar bar, + u64 cpu_addr, u32 aperture) +{ + struct cdns_pcie *pcie = &rc->pcie; + u32 addr0, addr1; + + addr0 = CDNS_PCIE_AT_IB_RP_BAR_ADDR0_NBITS(aperture) | + (lower_32_bits(cpu_addr) & GENMASK(31, 8)); + addr1 = upper_32_bits(cpu_addr); + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR0(bar), addr0); + cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR1(bar), addr1); +} + +static int cdns_pcie_host_bar_config(struct cdns_pcie_rc *rc, + struct resource_entry *entry, + enum cdns_pcie_rp_bar *index) +{ + u64 cpu_addr, pci_addr, size, winsize; + struct cdns_pcie *pcie = &rc->pcie; + struct device *dev = pcie->dev; + enum cdns_pcie_rp_bar bar; + unsigned long flags; + u32 aperture; + u32 value; + + cpu_addr = entry->res->start; + flags = entry->res->flags; + pci_addr = entry->res->start - entry->offset; + size = resource_size(entry->res); + bar = *index; + + if (entry->offset) { + dev_err(dev, "Cannot map PCI addr: %llx to CPU addr: %llx\n", + pci_addr, cpu_addr); + return -EINVAL; + } + + value = cdns_pcie_readl(pcie, CDNS_PCIE_LM_RC_BAR_CFG); + while (size > 0) { + if (bar > RP_NO_BAR) { + dev_err(dev, "Failed to map inbound regions!\n"); + return -EINVAL; + } + + winsize = size; + if (size > cdns_rp_bar_max_size[bar]) + winsize = cdns_rp_bar_max_size[bar]; + + aperture = ilog2(winsize); + + cdns_pcie_host_bar_ib_config(rc, bar, cpu_addr, aperture); + + if (bar == RP_NO_BAR) + break; + + if (winsize + cpu_addr >= SZ_4G) { + if (!(flags & IORESOURCE_PREFETCH)) + value |= LM_RC_BAR_CFG_CTRL_MEM_64BITS(bar); + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_64BITS(bar); + } else { + if (!(flags & IORESOURCE_PREFETCH)) + value |= LM_RC_BAR_CFG_CTRL_MEM_32BITS(bar); + value |= LM_RC_BAR_CFG_CTRL_PREF_MEM_32BITS(bar); + } + + value |= LM_RC_BAR_CFG_APERTURE(bar, aperture); + + size -= winsize; + cpu_addr += winsize; + bar++; + } + cdns_pcie_writel(pcie, CDNS_PCIE_LM_RC_BAR_CFG, value); + *index = bar; + + return 0; +} + +static int cdns_pcie_host_map_dma_ranges(struct cdns_pcie_rc *rc) +{ + enum cdns_pcie_rp_bar bar = RP_BAR0; + struct cdns_pcie *pcie = &rc->pcie; + struct device *dev = pcie->dev; + struct device_node *np = dev->of_node; + struct pci_host_bridge *bridge; + struct resource_entry *entry; + u32 no_bar_nbits = 32; + int err; + + bridge = pci_host_bridge_from_priv(rc); + if (!bridge) + return -ENOMEM; + + if (list_empty(&bridge->dma_ranges)) { + of_property_read_u32(np, "cdns,no-bar-match-nbits", + &no_bar_nbits); + cdns_pcie_host_bar_ib_config(rc, RP_NO_BAR, 0x0, no_bar_nbits); + return 0; + } + + resource_list_for_each_entry(entry, &bridge->dma_ranges) { + err = cdns_pcie_host_bar_config(rc, entry, &bar); + if (err) { + dev_err(dev, "Fail to configure IB using dma-ranges\n"); + return err; + } + } + + return 0; +} + static int cdns_pcie_host_init_address_translation(struct cdns_pcie_rc *rc) { struct cdns_pcie *pcie = &rc->pcie; @@ -160,16 +277,9 @@ static int cdns_pcie_host_init_address_translation(struct cdns_pcie_rc *rc) r++; } - /* - * Set Root Port no BAR match Inbound Translation registers: - * needed for MSI and DMA. - * Root Port BAR0 and BAR1 are disabled, hence no need to set their - * inbound translation registers. - */ - addr0 = CDNS_PCIE_AT_IB_RP_BAR_ADDR0_NBITS(rc->no_bar_nbits); - addr1 = 0; - cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR0(RP_NO_BAR), addr0); - cdns_pcie_writel(pcie, CDNS_PCIE_AT_IB_RP_BAR_ADDR1(RP_NO_BAR), addr1); + err = cdns_pcie_host_map_dma_ranges(rc); + if (err) + return err; return 0; } @@ -179,10 +289,16 @@ static int cdns_pcie_host_init(struct device *dev, struct cdns_pcie_rc *rc) { struct resource *bus_range = NULL; + struct pci_host_bridge *bridge; int err; + bridge = pci_host_bridge_from_priv(rc); + if (!bridge) + return -ENOMEM; + /* Parse our PCI ranges and request their resources */ - err = pci_parse_request_of_pci_ranges(dev, resources, NULL, &bus_range); + err = pci_parse_request_of_pci_ranges(dev, resources, + &bridge->dma_ranges, &bus_range); if (err) return err; @@ -239,9 +355,6 @@ int cdns_pcie_host_setup(struct cdns_pcie_rc *rc) pcie = &rc->pcie; pcie->is_rc = true; - rc->no_bar_nbits = 32; - of_property_read_u32(np, "cdns,no-bar-match-nbits", &rc->no_bar_nbits); - rc->vendor_id = 0xffff; of_property_read_u32(np, "vendor-id", &rc->vendor_id); diff --git a/drivers/pci/controller/cadence/pcie-cadence.h b/drivers/pci/controller/cadence/pcie-cadence.h index f349f5828a58..336237a7025c 100644 --- a/drivers/pci/controller/cadence/pcie-cadence.h +++ b/drivers/pci/controller/cadence/pcie-cadence.h @@ -92,6 +92,20 @@ #define CDNS_PCIE_LM_BAR_CFG_CTRL_MEM_64BITS 0x6 #define CDNS_PCIE_LM_BAR_CFG_CTRL_PREFETCH_MEM_64BITS 0x7 +#define LM_RC_BAR_CFG_CTRL_DISABLED(bar) \ + (CDNS_PCIE_LM_BAR_CFG_CTRL_DISABLED << (((bar) * 8) + 6)) +#define LM_RC_BAR_CFG_CTRL_IO_32BITS(bar) \ + (CDNS_PCIE_LM_BAR_CFG_CTRL_IO_32BITS << (((bar) * 8) + 6)) +#define LM_RC_BAR_CFG_CTRL_MEM_32BITS(bar) \ + (CDNS_PCIE_LM_BAR_CFG_CTRL_MEM_32BITS << (((bar) * 8) + 6)) +#define LM_RC_BAR_CFG_CTRL_PREF_MEM_32BITS(bar) \ + (CDNS_PCIE_LM_BAR_CFG_CTRL_PREFETCH_MEM_32BITS << (((bar) * 8) + 6)) +#define LM_RC_BAR_CFG_CTRL_MEM_64BITS(bar) \ + (CDNS_PCIE_LM_BAR_CFG_CTRL_MEM_64BITS << (((bar) * 8) + 6)) +#define LM_RC_BAR_CFG_CTRL_PREF_MEM_64BITS(bar) \ + (CDNS_PCIE_LM_BAR_CFG_CTRL_PREFETCH_MEM_64BITS << (((bar) * 8) + 6)) +#define LM_RC_BAR_CFG_APERTURE(bar, aperture) \ + (((aperture) - 2) << ((bar) * 8)) /* * Endpoint Function Registers (PCI configuration space for endpoint functions) @@ -266,8 +280,6 @@ struct cdns_pcie { * @bus_range: first/last buses behind the PCIe host controller * @cfg_base: IO mapped window to access the PCI configuration space of a * single function at a time - * @no_bar_nbits: Number of bits to keep for inbound (PCIe -> CPU) address - * translation (nbits sets into the "no BAR match" register) * @vendor_id: PCI vendor ID * @device_id: PCI device ID */ @@ -276,7 +288,6 @@ struct cdns_pcie_rc { struct resource *cfg_res; struct resource *bus_range; void __iomem *cfg_base; - u32 no_bar_nbits; u32 vendor_id; u32 device_id; };
Cadence PCIe core driver (host mode) uses "cdns,no-bar-match-nbits" property to configure the number of bits passed through from PCIe address to internal address in Inbound Address Translation register. This only used the NO MATCH BAR. However standard PCI dt-binding already defines "dma-ranges" to describe the address ranges accessible by PCIe controller. Add support in Cadence PCIe host driver to parse dma-ranges and configure the inbound regions for BAR0, BAR1 and NO MATCH BAR. Cadence IP specifies maximum size for BAR0 as 256GB, maximum size for BAR1 as 2 GB, so if the dma-ranges specifies a size larger than the maximum allowed, the driver will split and configure the BARs. Legacy device tree binding compatibility is maintained by retaining support for "cdns,no-bar-match-nbits". Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> --- .../controller/cadence/pcie-cadence-host.c | 141 ++++++++++++++++-- drivers/pci/controller/cadence/pcie-cadence.h | 17 ++- 2 files changed, 141 insertions(+), 17 deletions(-)