On Mon, 2020-03-02 at 18:51 +0100, Frederic Barrat wrote:
>
> Le 21/02/2020 à 04:27, Alastair D'Silva a écrit :
> > From: Alastair D'Silva <alastair@xxxxxxxxxxx>
> >
> > This patch reads timeouts & firmware version from the controller,
> > and
> > uses those timeouts to wait for the controller to report that it is
> > ready
> > before handing the memory over to libnvdimm.
> >
> > Signed-off-by: Alastair D'Silva <alastair@xxxxxxxxxxx>
> > ---
> > arch/powerpc/platforms/powernv/pmem/Makefile | 2 +-
> > arch/powerpc/platforms/powernv/pmem/ocxl.c | 92
> > +++++++++++++++++++
> > .../platforms/powernv/pmem/ocxl_internal.c | 19 ++++
> > .../platforms/powernv/pmem/ocxl_internal.h | 24 +++++
> > 4 files changed, 136 insertions(+), 1 deletion(-)
> > create mode 100644
> > arch/powerpc/platforms/powernv/pmem/ocxl_internal.c
> >
> > diff --git a/arch/powerpc/platforms/powernv/pmem/Makefile
> > b/arch/powerpc/platforms/powernv/pmem/Makefile
> > index 1c55c4193175..4ceda25907d4 100644
> > --- a/arch/powerpc/platforms/powernv/pmem/Makefile
> > +++ b/arch/powerpc/platforms/powernv/pmem/Makefile
> > @@ -4,4 +4,4 @@ ccflags-$(CONFIG_PPC_WERROR) += -Werror
> >
> > obj-$(CONFIG_OCXL_PMEM) += ocxlpmem.o
> >
> > -ocxlpmem-y := ocxl.o
> > +ocxlpmem-y := ocxl.o ocxl_internal.o
> > diff --git a/arch/powerpc/platforms/powernv/pmem/ocxl.c
> > b/arch/powerpc/platforms/powernv/pmem/ocxl.c
> > index 3c4eeb5dcc0f..431212c9f0cc 100644
> > --- a/arch/powerpc/platforms/powernv/pmem/ocxl.c
> > +++ b/arch/powerpc/platforms/powernv/pmem/ocxl.c
> > @@ -8,6 +8,7 @@
> >
> > #include <linux/module.h>
> > #include <misc/ocxl.h>
> > +#include <linux/delay.h>
> > #include <linux/ndctl.h>
> > #include <linux/mm_types.h>
> > #include <linux/memory_hotplug.h>
> > @@ -215,6 +216,36 @@ static int register_lpc_mem(struct ocxlpmem
> > *ocxlpmem)
> > return 0;
> > }
> >
> > +/**
> > + * is_usable() - Is a controller usable?
> > + * @ocxlpmem: the device metadata
> > + * @verbose: True to log errors
> > + * Return: true if the controller is usable
> > + */
> > +static bool is_usable(const struct ocxlpmem *ocxlpmem, bool
> > verbose)
> > +{
> > + u64 chi = 0;
> > + int rc = ocxlpmem_chi(ocxlpmem, &chi);
> > +
> > + if (rc < 0)
> > + return false;
> > +
> > + if (!(chi & GLOBAL_MMIO_CHI_CRDY)) {
> > + if (verbose)
> > + dev_err(&ocxlpmem->dev, "controller is not
> > ready.\n");
> > + return false;
> > + }
> > +
> > + if (!(chi & GLOBAL_MMIO_CHI_MA)) {
> > + if (verbose)
> > + dev_err(&ocxlpmem->dev,
> > + "controller does not have memory
> > available.\n");
> > + return false;
> > + }
> > +
> > + return true;
> > +}
> > +
> > /**
> > * allocate_minor() - Allocate a minor number to use for an
> > OpenCAPI pmem device
> > * @ocxlpmem: the device metadata
> > @@ -328,6 +359,48 @@ static void ocxlpmem_remove(struct pci_dev
> > *pdev)
> > }
> > }
> >
> > +/**
> > + * read_device_metadata() - Retrieve config information from the
> > AFU and save it for future use
> > + * @ocxlpmem: the device metadata
> > + * Return: 0 on success, negative on failure
> > + */
> > +static int read_device_metadata(struct ocxlpmem *ocxlpmem)
> > +{
> > + u64 val;
> > + int rc;
> > +
> > + rc = ocxl_global_mmio_read64(ocxlpmem->ocxl_afu,
> > GLOBAL_MMIO_CCAP0,
> > + OCXL_LITTLE_ENDIAN, &val);
> > + if (rc)
> > + return rc;
> > +
> > + ocxlpmem->scm_revision = val & 0xFFFF;
> > + ocxlpmem->read_latency = (val >> 32) & 0xFF;
> > + ocxlpmem->readiness_timeout = (val >> 48) & 0x0F;
> > + ocxlpmem->memory_available_timeout = val >> 52;
> > +
> > + rc = ocxl_global_mmio_read64(ocxlpmem->ocxl_afu,
> > GLOBAL_MMIO_CCAP1,
> > + OCXL_LITTLE_ENDIAN, &val);
> > + if (rc)
> > + return rc;
> > +
> > + ocxlpmem->max_controller_dump_size = val & 0xFFFFFFFF;
> > +
> > + // Extract firmware version text
> > + rc = ocxl_global_mmio_read64(ocxlpmem->ocxl_afu,
> > GLOBAL_MMIO_FWVER,
> > + OCXL_HOST_ENDIAN, (u64 *)ocxlpmem-
> > >fw_version);
> > + if (rc)
> > + return rc;
> > +
> > + ocxlpmem->fw_version[8] = '\0';
> > +
> > + dev_info(&ocxlpmem->dev,
> > + "Firmware version '%s' SCM revision %d:%d\n",
> > ocxlpmem->fw_version,
> > + ocxlpmem->scm_revision >> 4, ocxlpmem->scm_revision &
> > 0x0F);
> > +
> > + return 0;
> > +}
> > +
> > /**
> > * probe_function0() - Set up function 0 for an OpenCAPI
> > persistent memory device
> > * This is important as it enables templates higher than 0 across
> > all other functions,
> > @@ -368,6 +441,7 @@ static int probe(struct pci_dev *pdev, const
> > struct pci_device_id *ent)
> > {
> > struct ocxlpmem *ocxlpmem;
> > int rc;
> > + u16 elapsed, timeout;
> >
> > if (PCI_FUNC(pdev->devfn) == 0)
> > return probe_function0(pdev);
> > @@ -422,6 +496,24 @@ static int probe(struct pci_dev *pdev, const
> > struct pci_device_id *ent)
> > goto err;
> > }
> >
> > + if (read_device_metadata(ocxlpmem)) {
> > + dev_err(&pdev->dev, "Could not read metadata\n");
>
>
> Need to set rc
>
>
Whoops :)
>
> > + goto err;
> > + }
> > +
> > + elapsed = 0;
> > + timeout = ocxlpmem->readiness_timeout + ocxlpmem-
> > >memory_available_timeout;
> > + while (!is_usable(ocxlpmem, false)) {
> > + if (elapsed++ > timeout) {
> > + dev_warn(&ocxlpmem->dev, "OpenCAPI Persistent
> > Memory ready timeout.\n");
> > + (void)is_usable(ocxlpmem, true);
>
> I guess that extra call to is_usable() is just to log the cause of
> the
> error. However, with some bad luck, the call could now succeed.
>
Yeah, that's pretty ugly, I'll re-engineer it.
>
> Fred
>
>
> > + rc = -ENXIO;
> > + goto err;
> > + }
> > +
> > + msleep(1000);
> > + }
> > +
> > rc = register_lpc_mem(ocxlpmem);
> > if (rc) {
> > dev_err(&pdev->dev, "Could not register OpenCAPI
> > persistent memory with libnvdimm\n");
> > diff --git a/arch/powerpc/platforms/powernv/pmem/ocxl_internal.c
> > b/arch/powerpc/platforms/powernv/pmem/ocxl_internal.c
> > new file mode 100644
> > index 000000000000..617ca943b1b8
> > --- /dev/null
> > +++ b/arch/powerpc/platforms/powernv/pmem/ocxl_internal.c
> > @@ -0,0 +1,19 @@
> > +// SPDX-License-Identifier: GPL-2.0+
> > +// Copyright 2019 IBM Corp.
> > +
> > +#include <misc/ocxl.h>
> > +#include <linux/delay.h>
> > +#include "ocxl_internal.h"
> > +
> > +int ocxlpmem_chi(const struct ocxlpmem *ocxlpmem, u64 *chi)
> > +{
> > + u64 val;
> > + int rc = ocxl_global_mmio_read64(ocxlpmem->ocxl_afu,
> > GLOBAL_MMIO_CHI,
> > + OCXL_LITTLE_ENDIAN, &val);
> > + if (rc)
> > + return rc;
> > +
> > + *chi = val;
> > +
> > + return 0;
> > +}
> > diff --git a/arch/powerpc/platforms/powernv/pmem/ocxl_internal.h
> > b/arch/powerpc/platforms/powernv/pmem/ocxl_internal.h
> > index 9cf3e42750e7..ba0301533d00 100644
> > --- a/arch/powerpc/platforms/powernv/pmem/ocxl_internal.h
> > +++ b/arch/powerpc/platforms/powernv/pmem/ocxl_internal.h
> > @@ -97,4 +97,28 @@ struct ocxlpmem {
> > void *metadata_addr;
> > struct resource pmem_res;
> > struct nd_region *nd_region;
> > + char fw_version[8+1];
> > +
> > + u32 max_controller_dump_size;
> > + u16 scm_revision; // major/minor
> > + u8 readiness_timeout; /* The worst case time (in seconds) that
> > the host shall
> > + * wait for the controller to become
> > operational following a reset (CHI.CRDY).
> > + */
> > + u8 memory_available_timeout; /* The worst case time (in
> > seconds) that the host shall
> > + * wait for memory to become
> > available following a reset (CHI.MA).
> > + */
> > +
> > + u16 read_latency; /* The nominal measure of latency (in
> > nanoseconds)
> > + * associated with an unassisted read of a
> > memory block.
> > + * This represents the capability of the raw
> > media technology without assistance
> > + */
> > };
> > +
> > +/**
> > + * ocxlpmem_chi() - Get the value of the CHI register
> > + * @ocxlpmem: the device metadata
> > + * @chi: returns the CHI value
> > + *
> > + * Returns 0 on success, negative on error
> > + */
> > +int ocxlpmem_chi(const struct ocxlpmem *ocxlpmem, u64 *chi);
> >
--
Alastair D'Silva
Open Source Developer
Linux Technology Centre, IBM Australia
mob: 0423 762 819
