[Skiboot] [PATCH v2 2/7] skiboot: Find the IMA catalog

Oliver O'Halloran oohall at gmail.com
Tue Nov 22 12:16:55 AEDT 2016


On Fri, Nov 18, 2016 at 1:10 PM, Hemant Kumar <hemant at linux.vnet.ibm.com> wrote:
> IMA (In Memory Accumulation) catalog is a repository of information
> about the Performance Monitoring Units (PMUs) and their events under
> the IMA infrastructure. The information include :
>  - The PMU names
>  - Event names
>  - Event description
>  - Event offsets
>  - Event scale
>  - Event unit
>
> The catalog is provided as a flattened device tree (dtb). Processors
> with different PVR values may have different PMU or event names. Hence,
> for each processor, there can be multiple device tree binaries (dtbs)
> containing the IMA information. Each of the dtb is compressed and forms
> a sub-partition inside the PNOR partition "IMA_CATALOG". Here is a link
> to the commit adding this partition to PNOR :
> https://github.com/open-power/pnor/commit/c940142c6dc64dd176096dc648f433c889919e84
>
> So, each compressed dtb forms a sub-partition inside the IMA_CATALOG
> partition and can be accessed/loaded through a sub-partition id which
> is nothing but the PVR id. Based on the current processor's PVR, the
> appropriate sub-partion will be loaded.
>
>  IMA_CATALOG
>   partition
>  -----------   Sub-id (E.g)
>  |         |
>  |Catalog 1|    0x100
>  |---------|
>  |         |
>  |Catalog 2|    0x200    <------  Current processor's PVR (0x200)
>  |---------|
>  ...
>
> In the above example, if the current processor's PVR is 0x200, catalog 2
> should be loaded.

As far as I know a PNOR built for one system and/or processor revision
isn't going to be portable to another, so why are we doing this?

>
> Note however, that the catalog information is in the form of a dtb and
> the dtb is compressed too. So, the sub-partition loaded must be
> decompressed first before we can actually use it (which is done in
> subsequent patches).

You could add the decompression library before this patch and avoid
this awkward split entirely

>
> Signed-off-by: Hemant Kumar <hemant at linux.vnet.ibm.com>
> ---
> Chaneglog:
> v1 -> v2:
>  - Moved the macro checks for power8 to a separate function is_power8().
>
>  core/flash.c       |  1 +
>  core/init.c        |  4 ++++
>  hw/Makefile.inc    |  2 +-
>  hw/ima.c           | 69 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
>  include/ima.h      |  2 ++
>  include/platform.h |  1 +
>  6 files changed, 78 insertions(+), 1 deletion(-)
>  create mode 100644 hw/ima.c
>
> diff --git a/core/flash.c b/core/flash.c
> index 92de421..32b32dc 100644
> --- a/core/flash.c
> +++ b/core/flash.c
> @@ -416,6 +416,7 @@ static struct {
>         { RESOURCE_ID_KERNEL,   RESOURCE_SUBID_NONE,            "BOOTKERNEL" },
>         { RESOURCE_ID_INITRAMFS,RESOURCE_SUBID_NONE,            "ROOTFS" },
>         { RESOURCE_ID_CAPP,     RESOURCE_SUBID_SUPPORTED,       "CAPP" },
> +       { RESOURCE_ID_CATALOG,  RESOURCE_SUBID_SUPPORTED,       "IMA_CATALOG" },
>  };
>
>  /* This mimics the hostboot SBE format */
> diff --git a/core/init.c b/core/init.c
> index 43ce3a0..94a4708 100644
> --- a/core/init.c
> +++ b/core/init.c
> @@ -47,6 +47,7 @@
>  #include <nvram.h>
>  #include <libstb/stb.h>
>  #include <libstb/container.h>
> +#include <ima.h>
>
>  enum proc_gen proc_gen;
>
> @@ -903,6 +904,9 @@ void __noreturn __nomcount main_cpu_entry(const void *fdt)
>         /* Init SLW related stuff, including fastsleep */
>         slw_init();
>
> +       /* Init IMA related stuff (load the IMA dtb into memory) */
> +       ima_init();
> +
>         op_display(OP_LOG, OP_MOD_INIT, 0x0002);
>
>         /* Read in NVRAM and set it up */
> diff --git a/hw/Makefile.inc b/hw/Makefile.inc
> index a433c2b..8bcd203 100644
> --- a/hw/Makefile.inc
> +++ b/hw/Makefile.inc
> @@ -1,7 +1,7 @@
>  # -*-Makefile-*-
>  SUBDIRS += hw
>  HW_OBJS  = xscom.o chiptod.o gx.o cec.o lpc.o lpc-uart.o psi.o
> -HW_OBJS += homer.o slw.o occ.o fsi-master.o centaur.o
> +HW_OBJS += homer.o slw.o occ.o fsi-master.o centaur.o ima.o
>  HW_OBJS += nx.o nx-rng.o nx-crypto.o nx-842.o
>  HW_OBJS += p7ioc.o p7ioc-inits.o p7ioc-phb.o
>  HW_OBJS += phb3.o sfc-ctrl.o fake-rtc.o bt.o p8-i2c.o prd.o
> diff --git a/hw/ima.c b/hw/ima.c
> new file mode 100644
> index 0000000..7f9184f
> --- /dev/null
> +++ b/hw/ima.c
> @@ -0,0 +1,69 @@
> +/* Copyright 2016 IBM Corp.
> + *
> + * Licensed under the Apache License, Version 2.0 (the "License");
> + * you may not use this file except in compliance with the License.
> + * You may obtain a copy of the License at
> + *
> + *     http://www.apache.org/licenses/LICENSE-2.0
> + *
> + * Unless required by applicable law or agreed to in writing, software
> + * distributed under the License is distributed on an "AS IS" BASIS,
> + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
> + * implied.
> + * See the License for the specific language governing permissions and
> + * limitations under the License.
> + */
> +
> +#include <skiboot.h>
> +#include <xscom.h>
> +#include <ima.h>
> +#include <chip.h>
> +
> +static bool is_power8(void)
> +{
> +       struct proc_chip *chip;
> +
> +       chip = get_chip(this_cpu()->chip_id);
> +       if ((chip->type == PROC_CHIP_P8_MURANO) ||
> +           (chip->type == PROC_CHIP_P8_VENICE) ||
> +           (chip->type == PROC_CHIP_P8_NAPLES))
> +               return true;
> +       return false;
> +}

why not just use proc_gen?

> +
> +/*
> + * Fetch the IMA Catalog partition and find the appropriate sub-partition
> + * based on the platform's PVR.
> + */
> +void ima_init(void)
> +{
> +       char *buf = NULL;
> +       size_t size = IMA_DTB_SIZE;
> +       uint32_t pvr = mfspr(SPR_PVR);
> +       int ret;
> +
> +       /* Disable this for power 8 */
> +       if (is_power8())
> +               return;

Why are we disabling this for P8?   not supporting it on P8

Is it a P9 only feature? What about P7?

> +
> +       buf = malloc(IMA_DTB_SIZE);
> +       if (!buf) {
> +               prerror("IMA: unable to allocate memory\n");
> +               return;
> +       }
> +
> +       ret = start_preload_resource(RESOURCE_ID_CATALOG,
> +                                       pvr, buf, &size);
> +       if (ret != OPAL_SUCCESS)
> +               goto err;
> +
> +       ret = wait_for_resource_loaded(RESOURCE_ID_CATALOG,
> +                                         pvr);
> +       if (ret != OPAL_SUCCESS) {
> +               prerror("IMA: unable to load the catalog\n");
> +               goto err;
> +       }
> +
> +err:
> +       free(buf);
> +}



> diff --git a/include/ima.h b/include/ima.h
> index f628ee0..20f3c39 100644
> --- a/include/ima.h
> +++ b/include/ima.h
> @@ -109,4 +109,6 @@ struct ima_chip_cb
>  #define NEST_IMA_ENABLE                        0x1
>  #define NEST_IMA_DISABLE               0x2
>
> +void ima_init(void);
> +
>  #endif /* __IMA_H */
> diff --git a/include/platform.h b/include/platform.h
> index 334c0a4..02adb2f 100644
> --- a/include/platform.h
> +++ b/include/platform.h
> @@ -27,6 +27,7 @@ enum resource_id {
>         RESOURCE_ID_KERNEL,
>         RESOURCE_ID_INITRAMFS,
>         RESOURCE_ID_CAPP,
> +       RESOURCE_ID_CATALOG,
>  };
>  #define RESOURCE_SUBID_NONE 0
>  #define RESOURCE_SUBID_SUPPORTED 1
> --
> 2.7.4
>
> _______________________________________________
> Skiboot mailing list
> Skiboot at lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/skiboot


More information about the Skiboot mailing list