From: Chalapathi V <chalapathi.v@linux.ibm.com>
To: Nicholas Piggin <npiggin@gmail.com>, qemu-devel@nongnu.org
Cc: qemu-ppc@nongnu.org, fbarrat@linux.ibm.com, clg@kaod.org,
calebs@us.ibm.com, chalapathi.v@ibm.com,
saif.abrar@linux.vnet.ibm.com
Subject: Re: [PATCH v5 1/3] hw/ppc: Add pnv pervasive common chiplet units
Date: Sun, 26 Nov 2023 14:55:59 +0530 [thread overview]
Message-ID: <9295c310-8c99-4107-bf53-044bcf0c7488@linux.ibm.com> (raw)
In-Reply-To: <CX6ZWJOZ2QCQ.3LNMAZAVW61IJ@wheely>
On 24-11-2023 16:42, Nicholas Piggin wrote:
> On Fri Nov 24, 2023 at 8:15 PM AEST, Chalapathi V wrote:
>> This part of the patchset creates a common pervasive chiplet model where it
>> houses the common units of a chiplets.
>>
>> The chiplet control unit is common across chiplets and this commit implements
>> the pervasive chiplet model with chiplet control registers.
> This might be reworded to be a bit clearer, could provide a bit of
> background information too (in changelog or header comment):
>
> Status, configuration, and control of units in POWER chips is provided
> by the pervasive subsystem, which connects registers to the SCOM bus,
> which can be programmed by processor cores, other units on the chip,
> BMCs, or other POWER chips.
>
> A POWER10 chip is divided into logical pieces called chiplets. Chiplets
> are broadly divided into "core chiplets" (with the processor cores) and
> "nest chiplets" (with everything else). Each chiplet has an attachment
> to the pervasive bus (PIB) and with chiplet-specific registers. All nest
> chiplets have a common basic set of registers, which this model
> provides. They may also implement additional registers.
>
> That's my undertsanding, does it look right? If yes and this file is
> specifically for nest chiplets (not core), maybe the name should
> reflect that?
Sure Will change the name to PnvNestChipletPervasive and add some more
details to header comment.
Thank You,
Chalapathi
>> Signed-off-by: Chalapathi V <chalapathi.v@linux.ibm.com>
>> ---
>> include/hw/ppc/pnv_pervasive.h | 37 ++++++
>> include/hw/ppc/pnv_xscom.h | 3 +
>> hw/ppc/pnv_pervasive.c | 217 +++++++++++++++++++++++++++++++++
>> hw/ppc/meson.build | 1 +
>> 4 files changed, 258 insertions(+)
>> create mode 100644 include/hw/ppc/pnv_pervasive.h
>> create mode 100644 hw/ppc/pnv_pervasive.c
>>
>> diff --git a/include/hw/ppc/pnv_pervasive.h b/include/hw/ppc/pnv_pervasive.h
>> new file mode 100644
>> index 0000000000..d83f86df7b
>> --- /dev/null
>> +++ b/include/hw/ppc/pnv_pervasive.h
>> @@ -0,0 +1,37 @@
>> +/*
>> + * QEMU PowerPC pervasive common chiplet model
>> + *
>> + * Copyright (c) 2023, IBM Corporation.
>> + *
>> + * SPDX-License-Identifier: GPL-2.0-or-later
>> + *
>> + * This code is licensed under the GPL version 2 or later. See the
>> + * COPYING file in the top-level directory.
>> + */
>> +
>> +#ifndef PPC_PNV_PERVASIVE_H
>> +#define PPC_PNV_PERVASIVE_H
>> +
>> +#define TYPE_PNV_PERV "pnv-pervasive-chiplet"
>> +#define PNV_PERV(obj) OBJECT_CHECK(PnvPerv, (obj), TYPE_PNV_PERV)
>> +
>> +typedef struct PnvPervCtrlRegs {
>> +#define CPLT_CTRL_SIZE 6
>> + uint64_t cplt_ctrl[CPLT_CTRL_SIZE];
>> + uint64_t cplt_cfg0;
>> + uint64_t cplt_cfg1;
>> + uint64_t cplt_stat0;
>> + uint64_t cplt_mask0;
>> + uint64_t ctrl_protect_mode;
>> + uint64_t ctrl_atomic_lock;
>> +} PnvPervCtrlRegs;
>> +
>> +typedef struct PnvPerv {
> I don't want to bikeshed the name too much, but we have perv and
> pervasive, I'd prefer to use pervasive consistently.
>
> I would like the name to reflect that it is for common nest chiplet
> registers too, but maybe that's getting too ambitious...
>
> PnvNestChipletCommonRegs
> PnvNestChipletPervasive
>
> ?
Sure Will modify to PnvNestChipletPervasive.
>
>> + DeviceState parent;
>> + char *parent_obj_name;
>> + MemoryRegion xscom_perv_ctrl_regs;
>> + PnvPervCtrlRegs control_regs;
>> +} PnvPerv;
>> +
>> +void pnv_perv_dt(PnvPerv *perv, uint32_t base_addr, void *fdt, int offset);
>> +#endif /*PPC_PNV_PERVASIVE_H */
>> diff --git a/include/hw/ppc/pnv_xscom.h b/include/hw/ppc/pnv_xscom.h
>> index f5becbab41..d09d10f32b 100644
>> --- a/include/hw/ppc/pnv_xscom.h
>> +++ b/include/hw/ppc/pnv_xscom.h
>> @@ -170,6 +170,9 @@ struct PnvXScomInterfaceClass {
>> #define PNV10_XSCOM_XIVE2_BASE 0x2010800
>> #define PNV10_XSCOM_XIVE2_SIZE 0x400
>>
>> +#define PNV10_XSCOM_NEST1_CTRL_CHIPLET_BASE 0x3000000
>> +#define PNV10_XSCOM_CTRL_CHIPLET_SIZE 0x400
>> +
>> #define PNV10_XSCOM_PEC_NEST_BASE 0x3011800 /* index goes downwards ... */
>> #define PNV10_XSCOM_PEC_NEST_SIZE 0x100
>>
>> diff --git a/hw/ppc/pnv_pervasive.c b/hw/ppc/pnv_pervasive.c
>> new file mode 100644
>> index 0000000000..c925070798
>> --- /dev/null
>> +++ b/hw/ppc/pnv_pervasive.c
>> @@ -0,0 +1,217 @@
>> +/*
>> + * QEMU PowerPC pervasive common chiplet model
>> + *
>> + * Copyright (c) 2023, IBM Corporation.
>> + *
>> + * SPDX-License-Identifier: GPL-2.0-or-later
>> + *
>> + * This code is licensed under the GPL version 2 or later. See the
>> + * COPYING file in the top-level directory.
>> + */
>> +
>> +#include "qemu/osdep.h"
>> +#include "qemu/log.h"
>> +#include "hw/qdev-properties.h"
>> +#include "hw/ppc/pnv.h"
>> +#include "hw/ppc/pnv_xscom.h"
>> +#include "hw/ppc/pnv_pervasive.h"
>> +#include "hw/ppc/fdt.h"
>> +#include <libfdt.h>
>> +
>> +#define CPLT_CONF0 0x08
>> +#define CPLT_CONF0_OR 0x18
>> +#define CPLT_CONF0_CLEAR 0x28
>> +#define CPLT_CONF1 0x09
>> +#define CPLT_CONF1_OR 0x19
>> +#define CPLT_CONF1_CLEAR 0x29
>> +#define CPLT_STAT0 0x100
>> +#define CPLT_MASK0 0x101
>> +#define CPLT_PROTECT_MODE 0x3FE
>> +#define CPLT_ATOMIC_CLOCK 0x3FF
>> +
>> +static uint64_t pnv_chiplet_ctrl_read(void *opaque, hwaddr addr, unsigned size)
>> +{
>> + PnvPerv *perv = PNV_PERV(opaque);
>> + int reg = addr >> 3;
>> + uint64_t val = ~0ull;
>> +
>> + /* CPLT_CTRL0 to CPLT_CTRL5 */
>> + for (int i = 0; i < CPLT_CTRL_SIZE; i++) {
>> + if (reg == i) {
>> + return perv->control_regs.cplt_ctrl[i];
>> + } else if ((reg == (i + 0x10)) || (reg == (i + 0x20))) {
>> + qemu_log_mask(LOG_GUEST_ERROR, "%s: Write only register, ignoring "
>> + "xscom read at 0x%" PRIx64 "\n",
>> + __func__, (unsigned long)reg);
>> + return val;
>> + }
>> + }
>> +
>> + switch (reg) {
>> + case CPLT_CONF0:
>> + val = perv->control_regs.cplt_cfg0;
>> + break;
>> + case CPLT_CONF0_OR:
>> + case CPLT_CONF0_CLEAR:
>> + qemu_log_mask(LOG_GUEST_ERROR, "%s: Write only register, ignoring "
>> + "xscom read at 0x%" PRIx64 "\n",
>> + __func__, (unsigned long)reg);
>> + break;
>> + case CPLT_CONF1:
>> + val = perv->control_regs.cplt_cfg1;
>> + break;
>> + case CPLT_CONF1_OR:
>> + case CPLT_CONF1_CLEAR:
>> + qemu_log_mask(LOG_GUEST_ERROR, "%s: Write only register, ignoring "
>> + "xscom read at 0x%" PRIx64 "\n",
>> + __func__, (unsigned long)reg);
>> + break;
>> + case CPLT_STAT0:
>> + val = perv->control_regs.cplt_stat0;
>> + break;
>> + case CPLT_MASK0:
>> + val = perv->control_regs.cplt_mask0;
>> + break;
>> + case CPLT_PROTECT_MODE:
>> + val = perv->control_regs.ctrl_protect_mode;
>> + break;
>> + case CPLT_ATOMIC_CLOCK:
>> + val = perv->control_regs.ctrl_atomic_lock;
>> + break;
>> + default:
>> + qemu_log_mask(LOG_UNIMP, "%s: Chiplet_control_regs: Invalid xscom "
>> + "read at 0x%" PRIx64 "\n", __func__, (unsigned long)reg);
>> + }
>> + return val;
>> +}
>> +
>> +static void pnv_chiplet_ctrl_write(void *opaque, hwaddr addr,
>> + uint64_t val, unsigned size)
>> +{
>> + PnvPerv *perv = PNV_PERV(opaque);
>> + int reg = addr >> 3;
>> +
>> + /* CPLT_CTRL0 to CPLT_CTRL5 */
>> + for (int i = 0; i < CPLT_CTRL_SIZE; i++) {
>> + if (reg == i) {
>> + perv->control_regs.cplt_ctrl[i] = val;
>> + return;
>> + } else if (reg == (i + 0x10)) {
>> + perv->control_regs.cplt_ctrl[i] |= val;
>> + return;
>> + } else if (reg == (i + 0x20)) {
>> + perv->control_regs.cplt_ctrl[i] &= ~val;
>> + return;
>> + }
>> + }
>> +
>> + switch (reg) {
>> + case CPLT_CONF0:
>> + perv->control_regs.cplt_cfg0 = val;
>> + break;
>> + case CPLT_CONF0_OR:
>> + perv->control_regs.cplt_cfg0 |= val;
>> + break;
>> + case CPLT_CONF0_CLEAR:
>> + perv->control_regs.cplt_cfg0 &= ~val;
>> + break;
>> + case CPLT_CONF1:
>> + perv->control_regs.cplt_cfg1 = val;
>> + break;
>> + case CPLT_CONF1_OR:
>> + perv->control_regs.cplt_cfg1 |= val;
>> + break;
>> + case CPLT_CONF1_CLEAR:
>> + perv->control_regs.cplt_cfg1 &= ~val;
>> + break;
>> + case CPLT_STAT0:
>> + perv->control_regs.cplt_stat0 = val;
>> + break;
>> + case CPLT_MASK0:
>> + perv->control_regs.cplt_mask0 = val;
>> + break;
>> + case CPLT_PROTECT_MODE:
>> + perv->control_regs.ctrl_protect_mode = val;
>> + break;
>> + case CPLT_ATOMIC_CLOCK:
>> + perv->control_regs.ctrl_atomic_lock = val;
>> + break;
>> + default:
>> + qemu_log_mask(LOG_UNIMP, "%s: Chiplet_control_regs: Invalid xscom "
>> + "write at 0x%" PRIx64 "\n",
>> + __func__, (unsigned long)reg);
>> + }
>> +}
>> +
>> +static const MemoryRegionOps pnv_perv_control_xscom_ops = {
>> + .read = pnv_chiplet_ctrl_read,
>> + .write = pnv_chiplet_ctrl_write,
>> + .valid.min_access_size = 8,
>> + .valid.max_access_size = 8,
>> + .impl.min_access_size = 8,
>> + .impl.max_access_size = 8,
>> + .endianness = DEVICE_BIG_ENDIAN,
>> +};
>> +
>> +static void pnv_perv_realize(DeviceState *dev, Error **errp)
>> +{
>> + PnvPerv *perv = PNV_PERV(dev);
>> + g_autofree char *region_name = NULL;
>> + region_name = g_strdup_printf("xscom-%s-control-regs",
>> + perv->parent_obj_name);
>> +
>> + /* Chiplet control scoms */
>> + pnv_xscom_region_init(&perv->xscom_perv_ctrl_regs, OBJECT(perv),
>> + &pnv_perv_control_xscom_ops, perv, region_name,
>> + PNV10_XSCOM_CTRL_CHIPLET_SIZE);
>> +}
>> +
>> +void pnv_perv_dt(PnvPerv *perv, uint32_t base_addr, void *fdt, int offset)
>> +{
>> + g_autofree char *name = NULL;
>> + int perv_offset;
>> + const char compat[] = "ibm,power10-perv-chiplet";
> Should we provide this dt node?
>
> powernv boot environment is basically the end of hostboot boot but
> with device-tree instead of HDAT.
>
> The way OPAL (skiboot) works when booting hostboot is to first
> convert HDAT into device-tree. When booting QEMU it just uses QEMU
> device-tree directly.
>
> So adding new dt nodes is a bit strange. Normally if skiboot needs
> something new, it will extend the HDAT converter to find that
> thing and create a new device-tree entry from it. Then the QEMU
> model would add that same dt entry.
>
> So it's really skiboot that defines the device-tree IMO, and QEMU
> should just follow it. Unless there is good reason to do something
> else.
>
> Thanks,
> Nick
Thank You for the clarification. Will remove this node.
>> + uint32_t reg[] = {
>> + cpu_to_be32(base_addr),
>> + cpu_to_be32(PNV10_XSCOM_CTRL_CHIPLET_SIZE)
>> + };
>> +
>> + name = g_strdup_printf("%s-perv@%x", perv->parent_obj_name, base_addr);
>> + perv_offset = fdt_add_subnode(fdt, offset, name);
>> + _FDT(perv_offset);
>> +
>> + _FDT(fdt_setprop(fdt, perv_offset, "reg", reg, sizeof(reg)));
>> + _FDT(fdt_setprop(fdt, perv_offset, "compatible", compat, sizeof(compat)));
>> +}
>> +
>> +static Property pnv_perv_properties[] = {
>> + DEFINE_PROP_STRING("parent-obj-name", PnvPerv, parent_obj_name),
>> + DEFINE_PROP_END_OF_LIST(),
>> +};
>> +
>> +static void pnv_perv_class_init(ObjectClass *klass, void *data)
>> +{
>> + DeviceClass *dc = DEVICE_CLASS(klass);
>> +
>> + dc->desc = "PowerNV perv chiplet";
>> + dc->realize = pnv_perv_realize;
>> + device_class_set_props(dc, pnv_perv_properties);
>> +}
>> +
>> +static const TypeInfo pnv_perv_info = {
>> + .name = TYPE_PNV_PERV,
>> + .parent = TYPE_DEVICE,
>> + .instance_size = sizeof(PnvPerv),
>> + .class_init = pnv_perv_class_init,
>> + .interfaces = (InterfaceInfo[]) {
>> + { TYPE_PNV_XSCOM_INTERFACE },
>> + { }
>> + }
>> +};
>> +
>> +static void pnv_perv_register_types(void)
>> +{
>> + type_register_static(&pnv_perv_info);
>> +}
>> +
>> +type_init(pnv_perv_register_types);
>> diff --git a/hw/ppc/meson.build b/hw/ppc/meson.build
>> index ea44856d43..37a7a8935d 100644
>> --- a/hw/ppc/meson.build
>> +++ b/hw/ppc/meson.build
>> @@ -51,6 +51,7 @@ ppc_ss.add(when: 'CONFIG_POWERNV', if_true: files(
>> 'pnv_bmc.c',
>> 'pnv_homer.c',
>> 'pnv_pnor.c',
>> + 'pnv_pervasive.c',
>> ))
>> # PowerPC 4xx boards
>> ppc_ss.add(when: 'CONFIG_PPC405', if_true: files(
next prev parent reply other threads:[~2023-11-26 9:27 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-24 10:15 [PATCH v5 0/3] pnv nest1 chiplet model Chalapathi V
2023-11-24 10:15 ` [PATCH v5 1/3] hw/ppc: Add pnv pervasive common chiplet units Chalapathi V
2023-11-24 11:12 ` Nicholas Piggin
2023-11-26 9:25 ` Chalapathi V [this message]
2023-11-24 10:15 ` [PATCH v5 2/3] hw/ppc: Add nest1 chiplet model Chalapathi V
2023-11-24 11:26 ` Nicholas Piggin
2023-11-24 12:19 ` Cédric Le Goater
2023-11-24 13:01 ` Nicholas Piggin
2023-11-26 9:29 ` Chalapathi V
2023-11-24 10:15 ` [PATCH v5 3/3] hw/ppc: Nest1 chiplet wiring Chalapathi V
2023-11-24 11:28 ` Nicholas Piggin
2023-11-24 12:26 ` Cédric Le Goater
2023-11-24 12:47 ` Nicholas Piggin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=9295c310-8c99-4107-bf53-044bcf0c7488@linux.ibm.com \
--to=chalapathi.v@linux.ibm.com \
--cc=calebs@us.ibm.com \
--cc=chalapathi.v@ibm.com \
--cc=clg@kaod.org \
--cc=fbarrat@linux.ibm.com \
--cc=npiggin@gmail.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-ppc@nongnu.org \
--cc=saif.abrar@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).