Message ID | 1416413091-13452-4-git-send-email-tomasz.nowicki@linaro.org (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On Wed, Nov 19, 2014 at 05:04:48PM +0100, Tomasz Nowicki wrote: > We are going to use mmio_config_{} name convention across all architectures. mmio_config_*() are workarounds for an AMD Fam10h defect [1]. I would prefer not to extend these names to other architectures, because they should be able to use readb()/writeb()/etc. directly, as we did on x86 before 3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h"). In fact, it would be nice if we could use readb()/writeb()/etc. on x86, and only use mmio_config_*() when we're on AMD Fam10h. Maybe there could be a "struct pci_raw_ops pci_mmcfg_amd_fam10h" that used mmio_config_*(), and we could set raw_pci_ext_ops to those ops instead of the default ones when we're on AMD Fam10h. Then x86 should be able to use the generic readb()-based ops on most platforms. Bjorn [1] 3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h") Somewhere in this process, I'd like to update the AMD Fam10h comment to fix the typos and reference the spec that talks about this, i.e., the "BIOS and Kernel Developer's Guide (BKDG) For AMD Family 10h Processors," rev 3.48, sec 2.11.1, "MMIO Configuration Coding Requirements." That can be a separate patch since it's only cosmetic. > Currently it belongs to asm/pci_x86.h header which should be included > only for x86 specific files. From now on, those accessors are in asm/pci.h > header which can be included in non-architecture code much easier. > > Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org> > Tested-by: Hanjun Guo <hanjun.guo@linaro.org> > --- > arch/x86/include/asm/pci.h | 42 +++++++++++++++++++++++++++++++++++++++++ > arch/x86/include/asm/pci_x86.h | 43 ------------------------------------------ > 2 files changed, 42 insertions(+), 43 deletions(-) > > diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h > index 0892ea0..5ba3720 100644 > --- a/arch/x86/include/asm/pci.h > +++ b/arch/x86/include/asm/pci.h > @@ -71,6 +71,48 @@ void pcibios_set_master(struct pci_dev *dev); > struct irq_routing_table *pcibios_get_irq_routing_table(void); > int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); > > +/* > + * AMD Fam10h CPUs are buggy, and cannot access MMIO config space > + * on their northbrige except through the * %eax register. As such, you MUST > + * NOT use normal IOMEM accesses, you need to only use the magic mmio-config > + * accessor functions. > + * In fact just use pci_config_*, nothing else please. > + */ > +static inline unsigned char mmio_config_readb(void __iomem *pos) > +{ > + u8 val; > + asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); > + return val; > +} > + > +static inline unsigned short mmio_config_readw(void __iomem *pos) > +{ > + u16 val; > + asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); > + return val; > +} > + > +static inline unsigned int mmio_config_readl(void __iomem *pos) > +{ > + u32 val; > + asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); > + return val; > +} > + > +static inline void mmio_config_writeb(void __iomem *pos, u8 val) > +{ > + asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); > +} > + > +static inline void mmio_config_writew(void __iomem *pos, u16 val) > +{ > + asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); > +} > + > +static inline void mmio_config_writel(void __iomem *pos, u32 val) > +{ > + asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); > +} > > #define HAVE_PCI_MMAP > extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, > diff --git a/arch/x86/include/asm/pci_x86.h b/arch/x86/include/asm/pci_x86.h > index caba141..42e7332 100644 > --- a/arch/x86/include/asm/pci_x86.h > +++ b/arch/x86/include/asm/pci_x86.h > @@ -121,49 +121,6 @@ extern int __init pcibios_init(void); > extern int pci_legacy_init(void); > extern void pcibios_fixup_irqs(void); > > -/* > - * AMD Fam10h CPUs are buggy, and cannot access MMIO config space > - * on their northbrige except through the * %eax register. As such, you MUST > - * NOT use normal IOMEM accesses, you need to only use the magic mmio-config > - * accessor functions. > - * In fact just use pci_config_*, nothing else please. > - */ > -static inline unsigned char mmio_config_readb(void __iomem *pos) > -{ > - u8 val; > - asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); > - return val; > -} > - > -static inline unsigned short mmio_config_readw(void __iomem *pos) > -{ > - u16 val; > - asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); > - return val; > -} > - > -static inline unsigned int mmio_config_readl(void __iomem *pos) > -{ > - u32 val; > - asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); > - return val; > -} > - > -static inline void mmio_config_writeb(void __iomem *pos, u8 val) > -{ > - asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); > -} > - > -static inline void mmio_config_writew(void __iomem *pos, u16 val) > -{ > - asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); > -} > - > -static inline void mmio_config_writel(void __iomem *pos, u32 val) > -{ > - asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); > -} > - > #ifdef CONFIG_PCI > # ifdef CONFIG_ACPI > # define x86_default_pci_init pci_acpi_init > -- > 1.9.1 >
On 11.12.2014 00:17, Bjorn Helgaas wrote: > On Wed, Nov 19, 2014 at 05:04:48PM +0100, Tomasz Nowicki wrote: >> We are going to use mmio_config_{} name convention across all architectures. > > mmio_config_*() are workarounds for an AMD Fam10h defect [1]. I would prefer > not to extend these names to other architectures, because they should be > able to use readb()/writeb()/etc. directly, as we did on x86 before > 3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h"). > > In fact, it would be nice if we could use readb()/writeb()/etc. on x86, and > only use mmio_config_*() when we're on AMD Fam10h. Maybe there could be a > "struct pci_raw_ops pci_mmcfg_amd_fam10h" that used mmio_config_*(), and we > could set raw_pci_ext_ops to those ops instead of the default ones when > we're on AMD Fam10h. Then x86 should be able to use the generic > readb()-based ops on most platforms. That's good point, I will readb()/writeb() instead. > > Bjorn > > [1] 3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h") > > Somewhere in this process, I'd like to update the AMD Fam10h comment to fix > the typos and reference the spec that talks about this, i.e., the "BIOS and > Kernel Developer's Guide (BKDG) For AMD Family 10h Processors," rev 3.48, > sec 2.11.1, "MMIO Configuration Coding Requirements." > > That can be a separate patch since it's only cosmetic. Sure, will do. Regards, Tomasz > >> Currently it belongs to asm/pci_x86.h header which should be included >> only for x86 specific files. From now on, those accessors are in asm/pci.h >> header which can be included in non-architecture code much easier. >> >> Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org> >> Tested-by: Hanjun Guo <hanjun.guo@linaro.org> >> --- >> arch/x86/include/asm/pci.h | 42 +++++++++++++++++++++++++++++++++++++++++ >> arch/x86/include/asm/pci_x86.h | 43 ------------------------------------------ >> 2 files changed, 42 insertions(+), 43 deletions(-) >> >> diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h >> index 0892ea0..5ba3720 100644 >> --- a/arch/x86/include/asm/pci.h >> +++ b/arch/x86/include/asm/pci.h >> @@ -71,6 +71,48 @@ void pcibios_set_master(struct pci_dev *dev); >> struct irq_routing_table *pcibios_get_irq_routing_table(void); >> int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); >> >> +/* >> + * AMD Fam10h CPUs are buggy, and cannot access MMIO config space >> + * on their northbrige except through the * %eax register. As such, you MUST >> + * NOT use normal IOMEM accesses, you need to only use the magic mmio-config >> + * accessor functions. >> + * In fact just use pci_config_*, nothing else please. >> + */ >> +static inline unsigned char mmio_config_readb(void __iomem *pos) >> +{ >> + u8 val; >> + asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); >> + return val; >> +} >> + >> +static inline unsigned short mmio_config_readw(void __iomem *pos) >> +{ >> + u16 val; >> + asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); >> + return val; >> +} >> + >> +static inline unsigned int mmio_config_readl(void __iomem *pos) >> +{ >> + u32 val; >> + asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); >> + return val; >> +} >> + >> +static inline void mmio_config_writeb(void __iomem *pos, u8 val) >> +{ >> + asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); >> +} >> + >> +static inline void mmio_config_writew(void __iomem *pos, u16 val) >> +{ >> + asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> +} >> + >> +static inline void mmio_config_writel(void __iomem *pos, u32 val) >> +{ >> + asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> +} >> >> #define HAVE_PCI_MMAP >> extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, >> diff --git a/arch/x86/include/asm/pci_x86.h b/arch/x86/include/asm/pci_x86.h >> index caba141..42e7332 100644 >> --- a/arch/x86/include/asm/pci_x86.h >> +++ b/arch/x86/include/asm/pci_x86.h >> @@ -121,49 +121,6 @@ extern int __init pcibios_init(void); >> extern int pci_legacy_init(void); >> extern void pcibios_fixup_irqs(void); >> >> -/* >> - * AMD Fam10h CPUs are buggy, and cannot access MMIO config space >> - * on their northbrige except through the * %eax register. As such, you MUST >> - * NOT use normal IOMEM accesses, you need to only use the magic mmio-config >> - * accessor functions. >> - * In fact just use pci_config_*, nothing else please. >> - */ >> -static inline unsigned char mmio_config_readb(void __iomem *pos) >> -{ >> - u8 val; >> - asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); >> - return val; >> -} >> - >> -static inline unsigned short mmio_config_readw(void __iomem *pos) >> -{ >> - u16 val; >> - asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); >> - return val; >> -} >> - >> -static inline unsigned int mmio_config_readl(void __iomem *pos) >> -{ >> - u32 val; >> - asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); >> - return val; >> -} >> - >> -static inline void mmio_config_writeb(void __iomem *pos, u8 val) >> -{ >> - asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); >> -} >> - >> -static inline void mmio_config_writew(void __iomem *pos, u16 val) >> -{ >> - asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> -} >> - >> -static inline void mmio_config_writel(void __iomem *pos, u32 val) >> -{ >> - asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> -} >> - >> #ifdef CONFIG_PCI >> # ifdef CONFIG_ACPI >> # define x86_default_pci_init pci_acpi_init >> -- >> 1.9.1 >>
On 11.12.2014 00:17, Bjorn Helgaas wrote: > On Wed, Nov 19, 2014 at 05:04:48PM +0100, Tomasz Nowicki wrote: >> We are going to use mmio_config_{} name convention across all architectures. > > mmio_config_*() are workarounds for an AMD Fam10h defect [1]. I would prefer > not to extend these names to other architectures, because they should be > able to use readb()/writeb()/etc. directly, as we did on x86 before > 3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h"). > > In fact, it would be nice if we could use readb()/writeb()/etc. on x86, and > only use mmio_config_*() when we're on AMD Fam10h. Maybe there could be a > "struct pci_raw_ops pci_mmcfg_amd_fam10h" that used mmio_config_*(), and we > could set raw_pci_ext_ops to those ops instead of the default ones when > we're on AMD Fam10h. Then x86 should be able to use the generic > readb()-based ops on most platforms. While I do agree we should use readb()/writeb()... methods, I am not sure there is a nice way to use mmio_config_*() exclusively for AMD Fam10h. For x86, right now, we have three PCI config accessors sets: mmconfig_32.c, mmconfig_64.c, numachip.c and each are different. So one pci_mmcfg_amd_fam10h pci_raw_ops is not enough. I am thinking of having additional structure (integrated with "struct pci_mmcfg_region") that keeps MMIO accessors where readb()/writeb() would be the default one. For AMD Fam10h case we could tweak it as necessary. What do you thing Bjorn? Tomasz > > Bjorn > > [1] 3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h") > > Somewhere in this process, I'd like to update the AMD Fam10h comment to fix > the typos and reference the spec that talks about this, i.e., the "BIOS and > Kernel Developer's Guide (BKDG) For AMD Family 10h Processors," rev 3.48, > sec 2.11.1, "MMIO Configuration Coding Requirements." > > That can be a separate patch since it's only cosmetic. > >> Currently it belongs to asm/pci_x86.h header which should be included >> only for x86 specific files. From now on, those accessors are in asm/pci.h >> header which can be included in non-architecture code much easier. >> >> Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org> >> Tested-by: Hanjun Guo <hanjun.guo@linaro.org> >> --- >> arch/x86/include/asm/pci.h | 42 +++++++++++++++++++++++++++++++++++++++++ >> arch/x86/include/asm/pci_x86.h | 43 ------------------------------------------ >> 2 files changed, 42 insertions(+), 43 deletions(-) >> >> diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h >> index 0892ea0..5ba3720 100644 >> --- a/arch/x86/include/asm/pci.h >> +++ b/arch/x86/include/asm/pci.h >> @@ -71,6 +71,48 @@ void pcibios_set_master(struct pci_dev *dev); >> struct irq_routing_table *pcibios_get_irq_routing_table(void); >> int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); >> >> +/* >> + * AMD Fam10h CPUs are buggy, and cannot access MMIO config space >> + * on their northbrige except through the * %eax register. As such, you MUST >> + * NOT use normal IOMEM accesses, you need to only use the magic mmio-config >> + * accessor functions. >> + * In fact just use pci_config_*, nothing else please. >> + */ >> +static inline unsigned char mmio_config_readb(void __iomem *pos) >> +{ >> + u8 val; >> + asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); >> + return val; >> +} >> + >> +static inline unsigned short mmio_config_readw(void __iomem *pos) >> +{ >> + u16 val; >> + asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); >> + return val; >> +} >> + >> +static inline unsigned int mmio_config_readl(void __iomem *pos) >> +{ >> + u32 val; >> + asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); >> + return val; >> +} >> + >> +static inline void mmio_config_writeb(void __iomem *pos, u8 val) >> +{ >> + asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); >> +} >> + >> +static inline void mmio_config_writew(void __iomem *pos, u16 val) >> +{ >> + asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> +} >> + >> +static inline void mmio_config_writel(void __iomem *pos, u32 val) >> +{ >> + asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> +} >> >> #define HAVE_PCI_MMAP >> extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, >> diff --git a/arch/x86/include/asm/pci_x86.h b/arch/x86/include/asm/pci_x86.h >> index caba141..42e7332 100644 >> --- a/arch/x86/include/asm/pci_x86.h >> +++ b/arch/x86/include/asm/pci_x86.h >> @@ -121,49 +121,6 @@ extern int __init pcibios_init(void); >> extern int pci_legacy_init(void); >> extern void pcibios_fixup_irqs(void); >> >> -/* >> - * AMD Fam10h CPUs are buggy, and cannot access MMIO config space >> - * on their northbrige except through the * %eax register. As such, you MUST >> - * NOT use normal IOMEM accesses, you need to only use the magic mmio-config >> - * accessor functions. >> - * In fact just use pci_config_*, nothing else please. >> - */ >> -static inline unsigned char mmio_config_readb(void __iomem *pos) >> -{ >> - u8 val; >> - asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); >> - return val; >> -} >> - >> -static inline unsigned short mmio_config_readw(void __iomem *pos) >> -{ >> - u16 val; >> - asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); >> - return val; >> -} >> - >> -static inline unsigned int mmio_config_readl(void __iomem *pos) >> -{ >> - u32 val; >> - asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); >> - return val; >> -} >> - >> -static inline void mmio_config_writeb(void __iomem *pos, u8 val) >> -{ >> - asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); >> -} >> - >> -static inline void mmio_config_writew(void __iomem *pos, u16 val) >> -{ >> - asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> -} >> - >> -static inline void mmio_config_writel(void __iomem *pos, u32 val) >> -{ >> - asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); >> -} >> - >> #ifdef CONFIG_PCI >> # ifdef CONFIG_ACPI >> # define x86_default_pci_init pci_acpi_init >> -- >> 1.9.1 >>
On Tue, Feb 17, 2015 at 02:03:01PM +0100, Tomasz Nowicki wrote: > On 11.12.2014 00:17, Bjorn Helgaas wrote: > >On Wed, Nov 19, 2014 at 05:04:48PM +0100, Tomasz Nowicki wrote: > >>We are going to use mmio_config_{} name convention across all architectures. > > > >mmio_config_*() are workarounds for an AMD Fam10h defect [1]. I would prefer > >not to extend these names to other architectures, because they should be > >able to use readb()/writeb()/etc. directly, as we did on x86 before > >3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h"). > > > >In fact, it would be nice if we could use readb()/writeb()/etc. on x86, and > >only use mmio_config_*() when we're on AMD Fam10h. Maybe there could be a > >"struct pci_raw_ops pci_mmcfg_amd_fam10h" that used mmio_config_*(), and we > >could set raw_pci_ext_ops to those ops instead of the default ones when > >we're on AMD Fam10h. Then x86 should be able to use the generic > >readb()-based ops on most platforms. > > While I do agree we should use readb()/writeb()... methods, I am not > sure there is a nice way to use mmio_config_*() exclusively for AMD > Fam10h. For x86, right now, we have three PCI config accessors sets: > mmconfig_32.c, mmconfig_64.c, numachip.c and each are different. So > one pci_mmcfg_amd_fam10h pci_raw_ops is not enough. I am thinking of > having additional structure (integrated with "struct > pci_mmcfg_region") that keeps MMIO accessors where readb()/writeb() > would be the default one. For AMD Fam10h case we could tweak it as > necessary. What do you thing Bjorn? It's OK if we use mmio_config_*() for all x86 (not just AMD Fam10h); that's what we do today. It'd be *nicer* if we could use the workaround only for Fam10h, but if it complicates the code, don't bother. My main point is that I think your v1 posting requires every arch to implement mmio_config_*(), and they will all be the same. If an arch doesn't require a workaround, it shouldn't have to supply anything and we should default to readb/writeb/etc. > >[1] 3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h") > > > >Somewhere in this process, I'd like to update the AMD Fam10h comment to fix > >the typos and reference the spec that talks about this, i.e., the "BIOS and > >Kernel Developer's Guide (BKDG) For AMD Family 10h Processors," rev 3.48, > >sec 2.11.1, "MMIO Configuration Coding Requirements." > > > >That can be a separate patch since it's only cosmetic. > > > >>Currently it belongs to asm/pci_x86.h header which should be included > >>only for x86 specific files. From now on, those accessors are in asm/pci.h > >>header which can be included in non-architecture code much easier. > >> > >>Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org> > >>Tested-by: Hanjun Guo <hanjun.guo@linaro.org> > >>--- > >> arch/x86/include/asm/pci.h | 42 +++++++++++++++++++++++++++++++++++++++++ > >> arch/x86/include/asm/pci_x86.h | 43 ------------------------------------------ > >> 2 files changed, 42 insertions(+), 43 deletions(-) > >> > >>diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h > >>index 0892ea0..5ba3720 100644 > >>--- a/arch/x86/include/asm/pci.h > >>+++ b/arch/x86/include/asm/pci.h > >>@@ -71,6 +71,48 @@ void pcibios_set_master(struct pci_dev *dev); > >> struct irq_routing_table *pcibios_get_irq_routing_table(void); > >> int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); > >> > >>+/* > >>+ * AMD Fam10h CPUs are buggy, and cannot access MMIO config space > >>+ * on their northbrige except through the * %eax register. As such, you MUST > >>+ * NOT use normal IOMEM accesses, you need to only use the magic mmio-config > >>+ * accessor functions. > >>+ * In fact just use pci_config_*, nothing else please. > >>+ */ > >>+static inline unsigned char mmio_config_readb(void __iomem *pos) > >>+{ > >>+ u8 val; > >>+ asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); > >>+ return val; > >>+} > >>+ > >>+static inline unsigned short mmio_config_readw(void __iomem *pos) > >>+{ > >>+ u16 val; > >>+ asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); > >>+ return val; > >>+} > >>+ > >>+static inline unsigned int mmio_config_readl(void __iomem *pos) > >>+{ > >>+ u32 val; > >>+ asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); > >>+ return val; > >>+} > >>+ > >>+static inline void mmio_config_writeb(void __iomem *pos, u8 val) > >>+{ > >>+ asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); > >>+} > >>+ > >>+static inline void mmio_config_writew(void __iomem *pos, u16 val) > >>+{ > >>+ asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); > >>+} > >>+ > >>+static inline void mmio_config_writel(void __iomem *pos, u32 val) > >>+{ > >>+ asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); > >>+} > >> > >> #define HAVE_PCI_MMAP > >> extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, > >>diff --git a/arch/x86/include/asm/pci_x86.h b/arch/x86/include/asm/pci_x86.h > >>index caba141..42e7332 100644 > >>--- a/arch/x86/include/asm/pci_x86.h > >>+++ b/arch/x86/include/asm/pci_x86.h > >>@@ -121,49 +121,6 @@ extern int __init pcibios_init(void); > >> extern int pci_legacy_init(void); > >> extern void pcibios_fixup_irqs(void); > >> > >>-/* > >>- * AMD Fam10h CPUs are buggy, and cannot access MMIO config space > >>- * on their northbrige except through the * %eax register. As such, you MUST > >>- * NOT use normal IOMEM accesses, you need to only use the magic mmio-config > >>- * accessor functions. > >>- * In fact just use pci_config_*, nothing else please. > >>- */ > >>-static inline unsigned char mmio_config_readb(void __iomem *pos) > >>-{ > >>- u8 val; > >>- asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); > >>- return val; > >>-} > >>- > >>-static inline unsigned short mmio_config_readw(void __iomem *pos) > >>-{ > >>- u16 val; > >>- asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); > >>- return val; > >>-} > >>- > >>-static inline unsigned int mmio_config_readl(void __iomem *pos) > >>-{ > >>- u32 val; > >>- asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); > >>- return val; > >>-} > >>- > >>-static inline void mmio_config_writeb(void __iomem *pos, u8 val) > >>-{ > >>- asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); > >>-} > >>- > >>-static inline void mmio_config_writew(void __iomem *pos, u16 val) > >>-{ > >>- asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); > >>-} > >>- > >>-static inline void mmio_config_writel(void __iomem *pos, u32 val) > >>-{ > >>- asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); > >>-} > >>- > >> #ifdef CONFIG_PCI > >> # ifdef CONFIG_ACPI > >> # define x86_default_pci_init pci_acpi_init > >>-- > >>1.9.1 > >>
On Wed, Feb 18, 2015 at 12:27 PM, Bjorn Helgaas <bhelgaas@google.com> wrote: > On Tue, Feb 17, 2015 at 02:03:01PM +0100, Tomasz Nowicki wrote: >> On 11.12.2014 00:17, Bjorn Helgaas wrote: >> >On Wed, Nov 19, 2014 at 05:04:48PM +0100, Tomasz Nowicki wrote: >> >>We are going to use mmio_config_{} name convention across all architectures. >> > >> >mmio_config_*() are workarounds for an AMD Fam10h defect [1]. I would prefer >> >not to extend these names to other architectures, because they should be >> >able to use readb()/writeb()/etc. directly, as we did on x86 before >> >3320ad994afb ("x86: Work around mmio config space quirk on AMD Fam10h"). >> > >> >In fact, it would be nice if we could use readb()/writeb()/etc. on x86, and >> >only use mmio_config_*() when we're on AMD Fam10h. Maybe there could be a >> >"struct pci_raw_ops pci_mmcfg_amd_fam10h" that used mmio_config_*(), and we >> >could set raw_pci_ext_ops to those ops instead of the default ones when >> >we're on AMD Fam10h. Then x86 should be able to use the generic >> >readb()-based ops on most platforms. >> >> While I do agree we should use readb()/writeb()... methods, I am not >> sure there is a nice way to use mmio_config_*() exclusively for AMD >> Fam10h. For x86, right now, we have three PCI config accessors sets: >> mmconfig_32.c, mmconfig_64.c, numachip.c and each are different. So >> one pci_mmcfg_amd_fam10h pci_raw_ops is not enough. I am thinking of >> having additional structure (integrated with "struct >> pci_mmcfg_region") that keeps MMIO accessors where readb()/writeb() >> would be the default one. For AMD Fam10h case we could tweak it as >> necessary. What do you thing Bjorn? > > It's OK if we use mmio_config_*() for all x86 (not just AMD Fam10h); that's > what we do today. It'd be *nicer* if we could use the workaround only for > Fam10h, but if it complicates the code, don't bother. > > My main point is that I think your v1 posting requires every arch to > implement mmio_config_*(), and they will all be the same. If an arch > doesn't require a workaround, it shouldn't have to supply anything and we > should default to readb/writeb/etc. Perhaps the abstraction needs to move up a level to pci_ops functions. Then x86 can use the generic ones I added and AMD can use custom ones. BTW, there are some cases on MIPS that need special accessors. It's mainly a function of whether the accessors need to do endian swapping or not. Rob
diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h index 0892ea0..5ba3720 100644 --- a/arch/x86/include/asm/pci.h +++ b/arch/x86/include/asm/pci.h @@ -71,6 +71,48 @@ void pcibios_set_master(struct pci_dev *dev); struct irq_routing_table *pcibios_get_irq_routing_table(void); int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); +/* + * AMD Fam10h CPUs are buggy, and cannot access MMIO config space + * on their northbrige except through the * %eax register. As such, you MUST + * NOT use normal IOMEM accesses, you need to only use the magic mmio-config + * accessor functions. + * In fact just use pci_config_*, nothing else please. + */ +static inline unsigned char mmio_config_readb(void __iomem *pos) +{ + u8 val; + asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); + return val; +} + +static inline unsigned short mmio_config_readw(void __iomem *pos) +{ + u16 val; + asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); + return val; +} + +static inline unsigned int mmio_config_readl(void __iomem *pos) +{ + u32 val; + asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); + return val; +} + +static inline void mmio_config_writeb(void __iomem *pos, u8 val) +{ + asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); +} + +static inline void mmio_config_writew(void __iomem *pos, u16 val) +{ + asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); +} + +static inline void mmio_config_writel(void __iomem *pos, u32 val) +{ + asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); +} #define HAVE_PCI_MMAP extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, diff --git a/arch/x86/include/asm/pci_x86.h b/arch/x86/include/asm/pci_x86.h index caba141..42e7332 100644 --- a/arch/x86/include/asm/pci_x86.h +++ b/arch/x86/include/asm/pci_x86.h @@ -121,49 +121,6 @@ extern int __init pcibios_init(void); extern int pci_legacy_init(void); extern void pcibios_fixup_irqs(void); -/* - * AMD Fam10h CPUs are buggy, and cannot access MMIO config space - * on their northbrige except through the * %eax register. As such, you MUST - * NOT use normal IOMEM accesses, you need to only use the magic mmio-config - * accessor functions. - * In fact just use pci_config_*, nothing else please. - */ -static inline unsigned char mmio_config_readb(void __iomem *pos) -{ - u8 val; - asm volatile("movb (%1),%%al" : "=a" (val) : "r" (pos)); - return val; -} - -static inline unsigned short mmio_config_readw(void __iomem *pos) -{ - u16 val; - asm volatile("movw (%1),%%ax" : "=a" (val) : "r" (pos)); - return val; -} - -static inline unsigned int mmio_config_readl(void __iomem *pos) -{ - u32 val; - asm volatile("movl (%1),%%eax" : "=a" (val) : "r" (pos)); - return val; -} - -static inline void mmio_config_writeb(void __iomem *pos, u8 val) -{ - asm volatile("movb %%al,(%1)" : : "a" (val), "r" (pos) : "memory"); -} - -static inline void mmio_config_writew(void __iomem *pos, u16 val) -{ - asm volatile("movw %%ax,(%1)" : : "a" (val), "r" (pos) : "memory"); -} - -static inline void mmio_config_writel(void __iomem *pos, u32 val) -{ - asm volatile("movl %%eax,(%1)" : : "a" (val), "r" (pos) : "memory"); -} - #ifdef CONFIG_PCI # ifdef CONFIG_ACPI # define x86_default_pci_init pci_acpi_init