From: Jonathan Cameron via <qemu-devel@nongnu.org>
To: Fan Ni <nifan.cxl@gmail.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>, <linux-cxl@vger.kernel.org>,
<qemu-devel@nongnu.org>
Subject: Re: [PATCH -qemu] hw/cxl: Support get/set mctp response payload size
Date: Mon, 14 Oct 2024 12:10:13 +0100 [thread overview]
Message-ID: <20241014121013.000062c3@Huawei.com> (raw)
In-Reply-To: <Zwheg1hFMG_MYggX@fan>
On Thu, 10 Oct 2024 16:08:51 -0700
Fan Ni <nifan.cxl@gmail.com> wrote:
> On Wed, Oct 09, 2024 at 06:41:57PM -0700, Davidlohr Bueso wrote:
> > Add Get/Set Response Message Limit commands.
> >
> > Signed-off-by: Davidlohr Bueso <dave@stgolabs.net>
>
> The commit log may include the cxl spec reference. Otherwise,
>
> Reviewed-by: Fan Ni <fan.ni@samsung.com>
>
>
> > ---
> > hw/cxl/cxl-mailbox-utils.c | 68 ++++++++++++++++++++++++++++++++++++--
> > 1 file changed, 65 insertions(+), 3 deletions(-)
> >
> > diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
> > index c2d776bc96eb..98416af794bb 100644
> > --- a/hw/cxl/cxl-mailbox-utils.c
> > +++ b/hw/cxl/cxl-mailbox-utils.c
> > @@ -7,6 +7,8 @@
> > * COPYING file in the top-level directory.
> > */
> >
> > +#include <math.h>
> > +
> > #include "qemu/osdep.h"
> > #include "hw/pci/msi.h"
> > #include "hw/pci/msix.h"
> > @@ -56,6 +58,8 @@ enum {
> > INFOSTAT = 0x00,
> > #define IS_IDENTIFY 0x1
> > #define BACKGROUND_OPERATION_STATUS 0x2
> > + #define GET_RESPONSE_MSG_LIMIT 0x3
> > + #define SET_RESPONSE_MSG_LIMIT 0x4
> > EVENTS = 0x01,
> > #define GET_RECORDS 0x0
> > #define CLEAR_RECORDS 0x1
> > @@ -393,7 +397,7 @@ static CXLRetCode cmd_infostat_identify(const struct cxl_cmd *cmd,
> > uint16_t pcie_subsys_vid;
> > uint16_t pcie_subsys_id;
> > uint64_t sn;
> > - uint8_t max_message_size;
> > + uint8_t max_message_size;
Huh. I wonder how that misalignment got in.
I'll spin a separate tidy up patch to deal with that and
include it in a trivial series I'm sending later today.
> > uint8_t component_type;
> > } QEMU_PACKED *is_identify;
> > QEMU_BUILD_BUG_ON(sizeof(*is_identify) != 18);
> > @@ -422,12 +426,58 @@ static CXLRetCode cmd_infostat_identify(const struct cxl_cmd *cmd,
> > is_identify->component_type = 0x3; /* Type 3 */
> > }
> >
> > - /* TODO: Allow this to vary across different CCIs */
> > - is_identify->max_message_size = 9; /* 512 bytes - MCTP_CXL_MAILBOX_BYTES */
> > + is_identify->max_message_size = (uint8_t)log2(cci->payload_max);
> > *len_out = sizeof(*is_identify);
> > return CXL_MBOX_SUCCESS;
> > }
> >
> > +/* CXL r3.1 section 8.2.9.1.3: Get Response Message Limit (Opcode 0003h) */
> > +static CXLRetCode cmd_get_response_msg_limit(const struct cxl_cmd *cmd,
> > + uint8_t *payload_in,
> > + size_t len_in,
> > + uint8_t *payload_out,
> > + size_t *len_out,
> > + CXLCCI *cci)
> > +{
> > + struct {
> > + uint8_t rsp_limit;
> > + } QEMU_PACKED *get_rsp_msg_limit = (void *)payload_out;
> > + QEMU_BUILD_BUG_ON(sizeof(*get_rsp_msg_limit) != 1);
> > +
> > + get_rsp_msg_limit->rsp_limit = (uint8_t)log2(cci->payload_max);
> > +
> > + *len_out = sizeof(*get_rsp_msg_limit);
> > + return CXL_MBOX_SUCCESS;
> > +}
> > +
> > +/* CXL r3.1 section 8.2.9.1.4: Set Response Message Limit (Opcode 0004h) */
> > +static CXLRetCode cmd_set_response_msg_limit(const struct cxl_cmd *cmd,
> > + uint8_t *payload_in,
> > + size_t len_in,
> > + uint8_t *payload_out,
> > + size_t *len_out,
> > + CXLCCI *cci)
> > +{
> > + struct {
> > + uint8_t rsp_limit;
> > + } QEMU_PACKED *in = (void *)payload_in;
> > + QEMU_BUILD_BUG_ON(sizeof(*in) != 1);
> > + struct {
> > + uint8_t rsp_limit;
> > + } QEMU_PACKED *out = (void *)payload_out;
> > + QEMU_BUILD_BUG_ON(sizeof(*out) != 1);
> > +
> > + if (in->rsp_limit < 8 || in->rsp_limit > 10) {
Good to document why these values - possibly using defines.
I think 8 is the minimum the spec allows, but is the 10 based on
anything specific?
I'll carry this on my gitlab staging branch but want to
tidy this up before suggesting Michael picks it up.
I'll end up splitting this up a little though as only one of
the MCTP mailboxes on that tree is not dependent on the i2c mctp
stuff that is queued up behind Klaus' series. So I'll drag
the guts of this to near the top of my tree and include the extra
commands where that i2c_mctp is added.
hw/cxl/i2c_mctp_cxl: Initial device emulation
I'll push a new gitlab.com/jic23/qemu tree out later today with this
done.
Thanks,
Jonathan
> > + return CXL_MBOX_INVALID_INPUT;
> > + }
> > +
> > + cci->payload_max = 1 << in->rsp_limit;
> > + out->rsp_limit = in->rsp_limit;
> > +
> > + *len_out = sizeof(*out);
> > + return CXL_MBOX_SUCCESS;
> > +}
> > +
> > static void cxl_set_dsp_active_bm(PCIBus *b, PCIDevice *d,
> > void *private)
> > {
> > @@ -3000,6 +3050,10 @@ void cxl_initialize_mailbox_t3(CXLCCI *cci, DeviceState *d, size_t payload_max)
> >
> > static const struct cxl_cmd cxl_cmd_set_t3_mctp[256][256] = {
> > [INFOSTAT][IS_IDENTIFY] = { "IDENTIFY", cmd_infostat_identify, 0, 0 },
> > + [INFOSTAT][GET_RESPONSE_MSG_LIMIT] = { "GET_RESPONSE_MSG_LIMIT",
> > + cmd_get_response_msg_limit, 0, 0 },
> > + [INFOSTAT][SET_RESPONSE_MSG_LIMIT] = { "SET_RESPONSE_MSG_LIMIT",
> > + cmd_set_response_msg_limit, 1, 0 },
> > [TIMESTAMP][GET] = { "TIMESTAMP_GET", cmd_timestamp_get, 0, 0 },
> > [LOGS][GET_SUPPORTED] = { "LOGS_GET_SUPPORTED", cmd_logs_get_supported, 0,
> > 0 },
> > @@ -3035,6 +3089,10 @@ void cxl_initialize_t3_ld_cci(CXLCCI *cci, DeviceState *d, DeviceState *intf,
> >
> > static const struct cxl_cmd cxl_cmd_set_t3_fm_owned_ld_mctp[256][256] = {
> > [INFOSTAT][IS_IDENTIFY] = { "IDENTIFY", cmd_infostat_identify, 0, 0},
> > + [INFOSTAT][GET_RESPONSE_MSG_LIMIT] = { "GET_RESPONSE_MSG_LIMIT",
> > + cmd_get_response_msg_limit, 0, 0 },
> > + [INFOSTAT][SET_RESPONSE_MSG_LIMIT] = { "SET_RESPONSE_MSG_LIMIT",
> > + cmd_set_response_msg_limit, 1, 0 },
> > [LOGS][GET_SUPPORTED] = { "LOGS_GET_SUPPORTED", cmd_logs_get_supported, 0,
> > 0 },
> > [LOGS][GET_LOG] = { "LOGS_GET_LOG", cmd_logs_get_log, 0x18, 0 },
> > @@ -3055,6 +3113,10 @@ void cxl_initialize_t3_fm_owned_ld_mctpcci(CXLCCI *cci, DeviceState *d,
> >
> > static const struct cxl_cmd cxl_cmd_set_usp_mctp[256][256] = {
> > [INFOSTAT][IS_IDENTIFY] = { "IDENTIFY", cmd_infostat_identify, 0, 0 },
> > + [INFOSTAT][GET_RESPONSE_MSG_LIMIT] = { "GET_RESPONSE_MSG_LIMIT",
> > + cmd_get_response_msg_limit, 0, 0 },
> > + [INFOSTAT][SET_RESPONSE_MSG_LIMIT] = { "SET_RESPONSE_MSG_LIMIT",
> > + cmd_set_response_msg_limit, 1, 0 },
> > [LOGS][GET_SUPPORTED] = { "LOGS_GET_SUPPORTED", cmd_logs_get_supported,
> > 0, 0 },
> > [LOGS][GET_LOG] = { "LOGS_GET_LOG", cmd_logs_get_log, 0x18, 0 },
> > --
> > 2.46.1
> >
>
prev parent reply other threads:[~2024-10-14 11:11 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-10 1:41 [PATCH -qemu] hw/cxl: Support get/set mctp response payload size Davidlohr Bueso
2024-10-10 23:08 ` Fan Ni
2024-10-14 11:10 ` Jonathan Cameron via [this message]
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=20241014121013.000062c3@Huawei.com \
--to=qemu-devel@nongnu.org \
--cc=Jonathan.Cameron@Huawei.com \
--cc=dave@stgolabs.net \
--cc=linux-cxl@vger.kernel.org \
--cc=nifan.cxl@gmail.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).