linux/drivers/gpu/drm/i915/pxp/intel_pxp_gsccs.c

447 lines
13 KiB
C
Raw Permalink Normal View History

// SPDX-License-Identifier: MIT
/*
* Copyright(c) 2023 Intel Corporation.
*/
#include "gem/i915_gem_internal.h"
#include "gt/intel_context.h"
#include "gt/intel_gt.h"
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
#include "gt/uc/intel_gsc_fw.h"
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
#include "gt/uc/intel_gsc_uc_heci_cmd_submit.h"
#include "i915_drv.h"
#include "intel_pxp.h"
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
#include "intel_pxp_cmd_interface_42.h"
#include "intel_pxp_cmd_interface_43.h"
#include "intel_pxp_gsccs.h"
#include "intel_pxp_types.h"
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
static bool
drm/i915/pxp: Optimize GET_PARAM:PXP_STATUS After recent discussions with Mesa folks, it was requested that we optimize i915's GET_PARAM for the PXP_STATUS without changing the UAPI spec. Add these additional optimizations: - If any PXP initializatoin flow failed, then ensure that we catch it so that we can change the returned PXP_STATUS from "2" (i.e. 'PXP is supported but not yet ready') to "-ENODEV". This typically should not happen and if it does, we have a platform configuration issue. - If a PXP arbitration session creation event failed due to incorrect firmware version or blocking SOC fusing or blocking BIOS configuration (platform reasons that won't change if we retry), then reflect that blockage by also returning -ENODEV in the GET_PARAM:PXP_STATUS. - GET_PARAM:PXP_STATUS should not wait at all if PXP is supported but non-i915 dependencies (component-driver / firmware) we are still pending to complete the init flows. In this case, just return "2" immediately (i.e. 'PXP is supported but not yet ready'). Difference from prio revs: v3: - Rebase with latest tip that has pulled in setting the gsc fw load to fail if proxy init fails. v2: - Use a #define for the default readiness timeout (Vivaik). - Improve comments around the failing of proxy-init. v1: - Change the commit msg style to be imperative. (Jani) - Rename timeout to timeout_ms. (Jani) - Fix is_fw_err_platform_config to use higher order param (pxp) first. (Jani) Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Balasubrawmanian, Vivaik <vivaik.balasubrawmanian@intel.com> Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230802182550.1592926-1-alan.previn.teres.alexis@intel.com
2023-08-02 11:25:50 -07:00
is_fw_err_platform_config(struct intel_pxp *pxp, u32 type)
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
{
switch (type) {
case PXP_STATUS_ERROR_API_VERSION:
case PXP_STATUS_PLATFCONFIG_KF1_NOVERIF:
case PXP_STATUS_PLATFCONFIG_KF1_BAD:
drm/i915/pxp: Optimize GET_PARAM:PXP_STATUS After recent discussions with Mesa folks, it was requested that we optimize i915's GET_PARAM for the PXP_STATUS without changing the UAPI spec. Add these additional optimizations: - If any PXP initializatoin flow failed, then ensure that we catch it so that we can change the returned PXP_STATUS from "2" (i.e. 'PXP is supported but not yet ready') to "-ENODEV". This typically should not happen and if it does, we have a platform configuration issue. - If a PXP arbitration session creation event failed due to incorrect firmware version or blocking SOC fusing or blocking BIOS configuration (platform reasons that won't change if we retry), then reflect that blockage by also returning -ENODEV in the GET_PARAM:PXP_STATUS. - GET_PARAM:PXP_STATUS should not wait at all if PXP is supported but non-i915 dependencies (component-driver / firmware) we are still pending to complete the init flows. In this case, just return "2" immediately (i.e. 'PXP is supported but not yet ready'). Difference from prio revs: v3: - Rebase with latest tip that has pulled in setting the gsc fw load to fail if proxy init fails. v2: - Use a #define for the default readiness timeout (Vivaik). - Improve comments around the failing of proxy-init. v1: - Change the commit msg style to be imperative. (Jani) - Rename timeout to timeout_ms. (Jani) - Fix is_fw_err_platform_config to use higher order param (pxp) first. (Jani) Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Balasubrawmanian, Vivaik <vivaik.balasubrawmanian@intel.com> Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230802182550.1592926-1-alan.previn.teres.alexis@intel.com
2023-08-02 11:25:50 -07:00
pxp->platform_cfg_is_bad = true;
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
return true;
default:
break;
}
return false;
}
static const char *
fw_err_to_string(u32 type)
{
switch (type) {
case PXP_STATUS_ERROR_API_VERSION:
return "ERR_API_VERSION";
case PXP_STATUS_NOT_READY:
return "ERR_NOT_READY";
case PXP_STATUS_PLATFCONFIG_KF1_NOVERIF:
case PXP_STATUS_PLATFCONFIG_KF1_BAD:
return "ERR_PLATFORM_CONFIG";
default:
break;
}
return NULL;
}
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
static int
gsccs_send_message(struct intel_pxp *pxp,
void *msg_in, size_t msg_in_size,
void *msg_out, size_t msg_out_size_max,
size_t *msg_out_len,
u64 *gsc_msg_handle_retry)
{
struct intel_gt *gt = pxp->ctrl_gt;
struct drm_i915_private *i915 = gt->i915;
struct gsccs_session_resources *exec_res = &pxp->gsccs_res;
struct intel_gsc_mtl_header *header = exec_res->pkt_vaddr;
struct intel_gsc_heci_non_priv_pkt pkt;
size_t max_msg_size;
u32 reply_size;
int ret;
if (!exec_res->ce)
return -ENODEV;
max_msg_size = PXP43_MAX_HECI_INOUT_SIZE - sizeof(*header);
if (msg_in_size > max_msg_size || msg_out_size_max > max_msg_size)
return -ENOSPC;
if (!exec_res->pkt_vma || !exec_res->bb_vma)
return -ENOENT;
GEM_BUG_ON(exec_res->pkt_vma->size < (2 * PXP43_MAX_HECI_INOUT_SIZE));
mutex_lock(&pxp->tee_mutex);
memset(header, 0, sizeof(*header));
intel_gsc_uc_heci_cmd_emit_mtl_header(header, HECI_MEADDRESS_PXP,
msg_in_size + sizeof(*header),
exec_res->host_session_handle);
/* check if this is a host-session-handle cleanup call (empty packet) */
if (!msg_in && !msg_out)
header->flags |= GSC_INFLAG_MSG_CLEANUP;
/* copy caller provided gsc message handle if this is polling for a prior msg completion */
header->gsc_message_handle = *gsc_msg_handle_retry;
/* NOTE: zero size packets are used for session-cleanups */
if (msg_in && msg_in_size)
memcpy(exec_res->pkt_vaddr + sizeof(*header), msg_in, msg_in_size);
pkt.addr_in = i915_vma_offset(exec_res->pkt_vma);
pkt.size_in = header->message_size;
pkt.addr_out = pkt.addr_in + PXP43_MAX_HECI_INOUT_SIZE;
pkt.size_out = msg_out_size_max + sizeof(*header);
pkt.heci_pkt_vma = exec_res->pkt_vma;
pkt.bb_vma = exec_res->bb_vma;
/*
* Before submitting, let's clear-out the validity marker on the reply offset.
* We use offset PXP43_MAX_HECI_INOUT_SIZE for reply location so point header there.
*/
header = exec_res->pkt_vaddr + PXP43_MAX_HECI_INOUT_SIZE;
header->validity_marker = 0;
ret = intel_gsc_uc_heci_cmd_submit_nonpriv(&gt->uc.gsc,
exec_res->ce, &pkt, exec_res->bb_vaddr,
GSC_HECI_REPLY_LATENCY_MS);
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
if (ret) {
drm_err(&i915->drm, "failed to send gsc PXP msg (%d)\n", ret);
goto unlock;
}
/* Response validity marker, status and busyness */
if (header->validity_marker != GSC_HECI_VALIDITY_MARKER) {
drm_err(&i915->drm, "gsc PXP reply with invalid validity marker\n");
ret = -EINVAL;
goto unlock;
}
if (header->status != 0) {
drm_dbg(&i915->drm, "gsc PXP reply status has error = 0x%08x\n",
header->status);
ret = -EINVAL;
goto unlock;
}
if (header->flags & GSC_OUTFLAG_MSG_PENDING) {
drm_dbg(&i915->drm, "gsc PXP reply is busy\n");
/*
* When the GSC firmware replies with pending bit, it means that the requested
* operation has begun but the completion is pending and the caller needs
* to re-request with the gsc_message_handle that was returned by the firmware.
* until the pending bit is turned off.
*/
*gsc_msg_handle_retry = header->gsc_message_handle;
ret = -EAGAIN;
goto unlock;
}
reply_size = header->message_size - sizeof(*header);
if (reply_size > msg_out_size_max) {
drm/i915/pxp: Fix size_t format specifier in gsccs_send_message() When building ARCH=i386 allmodconfig, the following warning occurs: In file included from include/linux/device.h:15, from include/linux/node.h:18, from include/linux/cpu.h:17, from include/linux/static_call.h:135, from arch/x86/include/asm/perf_event.h:5, from include/linux/perf_event.h:25, from drivers/gpu/drm/i915/i915_pmu.h:11, from drivers/gpu/drm/i915/gt/intel_engine_types.h:21, from drivers/gpu/drm/i915/gt/intel_context_types.h:18, from drivers/gpu/drm/i915/gem/i915_gem_context_types.h:20, from drivers/gpu/drm/i915/i915_request.h:34, from drivers/gpu/drm/i915/i915_active.h:13, from drivers/gpu/drm/i915/gt/intel_context.h:13, from drivers/gpu/drm/i915/pxp/intel_pxp_gsccs.c:8: drivers/gpu/drm/i915/pxp/intel_pxp_gsccs.c: In function 'gsccs_send_message': include/drm/drm_print.h:456:39: error: format '%ld' expects argument of type 'long int', but argument 4 has type 'size_t' {aka 'unsigned int'} [-Werror=format=] 456 | dev_##level##type((drm)->dev, "[drm] " fmt, ##__VA_ARGS__) | ^~~~~~~~ include/linux/dev_printk.h:110:30: note: in definition of macro 'dev_printk_index_wrap' 110 | _p_func(dev, fmt, ##__VA_ARGS__); \ | ^~~ include/linux/dev_printk.h:146:61: note: in expansion of macro 'dev_fmt' 146 | dev_printk_index_wrap(_dev_warn, KERN_WARNING, dev, dev_fmt(fmt), ##__VA_ARGS__) | ^~~~~~~ include/drm/drm_print.h:456:9: note: in expansion of macro 'dev_warn' 456 | dev_##level##type((drm)->dev, "[drm] " fmt, ##__VA_ARGS__) | ^~~~ include/drm/drm_print.h:466:9: note: in expansion of macro '__drm_printk' 466 | __drm_printk((drm), warn,, fmt, ##__VA_ARGS__) | ^~~~~~~~~~~~ drivers/gpu/drm/i915/pxp/intel_pxp_gsccs.c:146:17: note: in expansion of macro 'drm_warn' 146 | drm_warn(&i915->drm, "caller with insufficient PXP reply size %u (%ld)\n", | ^~~~~~~~ cc1: all warnings being treated as errors Use the '%zu' format specifier, as the variable is a 'size_t'. Fixes: dc9ac125d81f ("drm/i915/pxp: Add GSC-CS backend to send GSC fw messages") Signed-off-by: Nathan Chancellor <nathan@kernel.org> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230530-i915-pxp-size_t-wformat-v1-1-9631081e2e5b@kernel.org
2023-05-30 11:37:56 -07:00
drm_warn(&i915->drm, "caller with insufficient PXP reply size %u (%zu)\n",
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
reply_size, msg_out_size_max);
reply_size = msg_out_size_max;
}
if (msg_out)
memcpy(msg_out, exec_res->pkt_vaddr + PXP43_MAX_HECI_INOUT_SIZE + sizeof(*header),
reply_size);
if (msg_out_len)
*msg_out_len = reply_size;
unlock:
mutex_unlock(&pxp->tee_mutex);
return ret;
}
static int
gsccs_send_message_retry_complete(struct intel_pxp *pxp,
void *msg_in, size_t msg_in_size,
void *msg_out, size_t msg_out_size_max,
size_t *msg_out_len)
{
u64 gsc_session_retry = 0;
int ret, tries = 0;
/*
* Keep sending request if GSC firmware was busy. Based on fw specs +
* sw overhead (and testing) we expect a worst case pending-bit delay of
* GSC_PENDING_RETRY_MAXCOUNT x GSC_PENDING_RETRY_PAUSE_MS millisecs.
*/
do {
ret = gsccs_send_message(pxp, msg_in, msg_in_size, msg_out, msg_out_size_max,
msg_out_len, &gsc_session_retry);
/* Only try again if gsc says so */
if (ret != -EAGAIN)
break;
msleep(GSC_PENDING_RETRY_PAUSE_MS);
} while (++tries < GSC_PENDING_RETRY_MAXCOUNT);
return ret;
}
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
bool intel_pxp_gsccs_is_ready_for_sessions(struct intel_pxp *pxp)
{
/*
* GSC-fw loading, HuC-fw loading, HuC-fw authentication and
* GSC-proxy init flow (requiring an mei component driver)
* must all occur first before we can start requesting for PXP
* sessions. Checking for completion on HuC authentication and
* gsc-proxy init flow (the last set of dependencies that
* are out of order) will suffice.
*/
if (intel_huc_is_authenticated(&pxp->ctrl_gt->uc.huc, INTEL_HUC_AUTH_BY_GSC) &&
2023-06-15 14:19:40 -07:00
intel_gsc_uc_fw_proxy_init_done(&pxp->ctrl_gt->uc.gsc, true))
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
return true;
return false;
}
int intel_pxp_gsccs_create_session(struct intel_pxp *pxp,
int arb_session_id)
{
struct drm_i915_private *i915 = pxp->ctrl_gt->i915;
struct pxp43_create_arb_in msg_in = {};
struct pxp43_create_arb_out msg_out = {};
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
int ret;
msg_in.header.api_version = PXP_APIVER(4, 3);
msg_in.header.command_id = PXP43_CMDID_INIT_SESSION;
msg_in.header.stream_id = (FIELD_PREP(PXP43_INIT_SESSION_APPID, arb_session_id) |
FIELD_PREP(PXP43_INIT_SESSION_VALID, 1) |
FIELD_PREP(PXP43_INIT_SESSION_APPTYPE, 0));
msg_in.header.buffer_len = sizeof(msg_in) - sizeof(msg_in.header);
msg_in.protection_mode = PXP43_INIT_SESSION_PROTECTION_ARB;
ret = gsccs_send_message_retry_complete(pxp,
&msg_in, sizeof(msg_in),
&msg_out, sizeof(msg_out), NULL);
if (ret) {
drm_err(&i915->drm, "Failed to init session %d, ret=[%d]\n", arb_session_id, ret);
} else if (msg_out.header.status != 0) {
drm/i915/pxp: Optimize GET_PARAM:PXP_STATUS After recent discussions with Mesa folks, it was requested that we optimize i915's GET_PARAM for the PXP_STATUS without changing the UAPI spec. Add these additional optimizations: - If any PXP initializatoin flow failed, then ensure that we catch it so that we can change the returned PXP_STATUS from "2" (i.e. 'PXP is supported but not yet ready') to "-ENODEV". This typically should not happen and if it does, we have a platform configuration issue. - If a PXP arbitration session creation event failed due to incorrect firmware version or blocking SOC fusing or blocking BIOS configuration (platform reasons that won't change if we retry), then reflect that blockage by also returning -ENODEV in the GET_PARAM:PXP_STATUS. - GET_PARAM:PXP_STATUS should not wait at all if PXP is supported but non-i915 dependencies (component-driver / firmware) we are still pending to complete the init flows. In this case, just return "2" immediately (i.e. 'PXP is supported but not yet ready'). Difference from prio revs: v3: - Rebase with latest tip that has pulled in setting the gsc fw load to fail if proxy init fails. v2: - Use a #define for the default readiness timeout (Vivaik). - Improve comments around the failing of proxy-init. v1: - Change the commit msg style to be imperative. (Jani) - Rename timeout to timeout_ms. (Jani) - Fix is_fw_err_platform_config to use higher order param (pxp) first. (Jani) Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Balasubrawmanian, Vivaik <vivaik.balasubrawmanian@intel.com> Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230802182550.1592926-1-alan.previn.teres.alexis@intel.com
2023-08-02 11:25:50 -07:00
if (is_fw_err_platform_config(pxp, msg_out.header.status)) {
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
drm_info_once(&i915->drm,
"PXP init-session-%d failed due to BIOS/SOC:0x%08x:%s\n",
arb_session_id, msg_out.header.status,
fw_err_to_string(msg_out.header.status));
} else {
drm_dbg(&i915->drm, "PXP init-session-%d failed 0x%08x:%st:\n",
arb_session_id, msg_out.header.status,
fw_err_to_string(msg_out.header.status));
drm_dbg(&i915->drm, " cmd-detail: ID=[0x%08x],API-Ver-[0x%08x]\n",
msg_in.header.command_id, msg_in.header.api_version);
}
}
return ret;
}
void intel_pxp_gsccs_end_arb_fw_session(struct intel_pxp *pxp, u32 session_id)
{
struct drm_i915_private *i915 = pxp->ctrl_gt->i915;
struct pxp42_inv_stream_key_in msg_in = {};
struct pxp42_inv_stream_key_out msg_out = {};
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
int ret = 0;
/*
* Stream key invalidation reuses the same version 4.2 input/output
* command format but firmware requires 4.3 API interaction
*/
msg_in.header.api_version = PXP_APIVER(4, 3);
msg_in.header.command_id = PXP42_CMDID_INVALIDATE_STREAM_KEY;
msg_in.header.buffer_len = sizeof(msg_in) - sizeof(msg_in.header);
msg_in.header.stream_id = FIELD_PREP(PXP_CMDHDR_EXTDATA_SESSION_VALID, 1);
msg_in.header.stream_id |= FIELD_PREP(PXP_CMDHDR_EXTDATA_APP_TYPE, 0);
msg_in.header.stream_id |= FIELD_PREP(PXP_CMDHDR_EXTDATA_SESSION_ID, session_id);
ret = gsccs_send_message_retry_complete(pxp,
&msg_in, sizeof(msg_in),
&msg_out, sizeof(msg_out), NULL);
if (ret) {
drm_err(&i915->drm, "Failed to inv-stream-key-%u, ret=[%d]\n",
session_id, ret);
} else if (msg_out.header.status != 0) {
drm/i915/pxp: Optimize GET_PARAM:PXP_STATUS After recent discussions with Mesa folks, it was requested that we optimize i915's GET_PARAM for the PXP_STATUS without changing the UAPI spec. Add these additional optimizations: - If any PXP initializatoin flow failed, then ensure that we catch it so that we can change the returned PXP_STATUS from "2" (i.e. 'PXP is supported but not yet ready') to "-ENODEV". This typically should not happen and if it does, we have a platform configuration issue. - If a PXP arbitration session creation event failed due to incorrect firmware version or blocking SOC fusing or blocking BIOS configuration (platform reasons that won't change if we retry), then reflect that blockage by also returning -ENODEV in the GET_PARAM:PXP_STATUS. - GET_PARAM:PXP_STATUS should not wait at all if PXP is supported but non-i915 dependencies (component-driver / firmware) we are still pending to complete the init flows. In this case, just return "2" immediately (i.e. 'PXP is supported but not yet ready'). Difference from prio revs: v3: - Rebase with latest tip that has pulled in setting the gsc fw load to fail if proxy init fails. v2: - Use a #define for the default readiness timeout (Vivaik). - Improve comments around the failing of proxy-init. v1: - Change the commit msg style to be imperative. (Jani) - Rename timeout to timeout_ms. (Jani) - Fix is_fw_err_platform_config to use higher order param (pxp) first. (Jani) Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Balasubrawmanian, Vivaik <vivaik.balasubrawmanian@intel.com> Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230802182550.1592926-1-alan.previn.teres.alexis@intel.com
2023-08-02 11:25:50 -07:00
if (is_fw_err_platform_config(pxp, msg_out.header.status)) {
drm/i915/pxp: Add ARB session creation and cleanup Add MTL's function for ARB session creation using PXP firmware version 4.3 ABI structure format. While relooking at the ARB session creation flow in intel_pxp_start, let's address missing UAPI documentation. Without actually changing backward compatible behavior, update i915's drm-uapi comments that describe the possible error values when creating a context with I915_CONTEXT_PARAM_PROTECTED_CONTENT: Since the first merge of PXP support on ADL, i915 returns -ENXIO if a dependency such as firmware or component driver was yet to be loaded or returns -EIO if the creation attempt failed when requested by the PXP firmware (specific firmware error responses are reported in dmesg). Add MTL's function for ARB session invalidation but this reuses PXP firmware version 4.2 ABI structure format. For both cases, in the back-end gsccs functions for sending messages to the firmware inspect the GSC-CS-Mem-Header's pending-bit which means the GSC firmware is busy and we should retry. Given the last hw requirement, lets also update functions in front-end layer that wait for session creation or teardown completion to use new worst case timeout periods. Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-6-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:35 -07:00
drm_info_once(&i915->drm,
"PXP inv-stream-key-%u failed due to BIOS/SOC :0x%08x:%s\n",
session_id, msg_out.header.status,
fw_err_to_string(msg_out.header.status));
} else {
drm_dbg(&i915->drm, "PXP inv-stream-key-%u failed 0x%08x:%s:\n",
session_id, msg_out.header.status,
fw_err_to_string(msg_out.header.status));
drm_dbg(&i915->drm, " cmd-detail: ID=[0x%08x],API-Ver-[0x%08x]\n",
msg_in.header.command_id, msg_in.header.api_version);
}
}
}
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
static void
gsccs_cleanup_fw_host_session_handle(struct intel_pxp *pxp)
{
struct drm_i915_private *i915 = pxp->ctrl_gt->i915;
int ret;
ret = gsccs_send_message_retry_complete(pxp, NULL, 0, NULL, 0, NULL);
if (ret)
drm_dbg(&i915->drm, "Failed to send gsccs msg host-session-cleanup: ret=[%d]\n",
ret);
}
static void
gsccs_destroy_execution_resource(struct intel_pxp *pxp)
{
struct gsccs_session_resources *exec_res = &pxp->gsccs_res;
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
if (exec_res->host_session_handle)
gsccs_cleanup_fw_host_session_handle(pxp);
if (exec_res->ce)
intel_context_put(exec_res->ce);
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
if (exec_res->bb_vma)
i915_vma_unpin_and_release(&exec_res->bb_vma, I915_VMA_RELEASE_MAP);
if (exec_res->pkt_vma)
i915_vma_unpin_and_release(&exec_res->pkt_vma, I915_VMA_RELEASE_MAP);
memset(exec_res, 0, sizeof(*exec_res));
}
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
static int
gsccs_create_buffer(struct intel_gt *gt,
const char *bufname, size_t size,
struct i915_vma **vma, void **map)
{
struct drm_i915_private *i915 = gt->i915;
struct drm_i915_gem_object *obj;
int err = 0;
obj = i915_gem_object_create_internal(i915, size);
if (IS_ERR(obj)) {
drm_err(&i915->drm, "Failed to allocate gsccs backend %s.\n", bufname);
err = PTR_ERR(obj);
goto out_none;
}
*vma = i915_vma_instance(obj, gt->vm, NULL);
if (IS_ERR(*vma)) {
drm_err(&i915->drm, "Failed to vma-instance gsccs backend %s.\n", bufname);
err = PTR_ERR(*vma);
goto out_put;
}
/* return a virtual pointer */
*map = i915_gem_object_pin_map_unlocked(obj, intel_gt_coherent_map_type(gt, obj, true));
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
if (IS_ERR(*map)) {
drm_err(&i915->drm, "Failed to map gsccs backend %s.\n", bufname);
err = PTR_ERR(*map);
goto out_put;
}
/* all PXP sessions commands are treated as non-privileged */
err = i915_vma_pin(*vma, 0, 0, PIN_USER);
if (err) {
drm_err(&i915->drm, "Failed to vma-pin gsccs backend %s.\n", bufname);
goto out_unmap;
}
return 0;
out_unmap:
i915_gem_object_unpin_map(obj);
out_put:
i915_gem_object_put(obj);
out_none:
*vma = NULL;
*map = NULL;
return err;
}
static int
gsccs_allocate_execution_resource(struct intel_pxp *pxp)
{
struct intel_gt *gt = pxp->ctrl_gt;
struct gsccs_session_resources *exec_res = &pxp->gsccs_res;
struct intel_engine_cs *engine = gt->engine[GSC0];
struct intel_context *ce;
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
int err = 0;
/*
* First, ensure the GSC engine is present.
* NOTE: Backend would only be called with the correct gt.
*/
if (!engine)
return -ENODEV;
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
/*
* Now, allocate, pin and map two objects, one for the heci message packet
* and another for the batch buffer we submit into GSC engine (that includes the packet).
* NOTE: GSC-CS backend is currently only supported on MTL, so we allocate shmem.
*/
err = gsccs_create_buffer(pxp->ctrl_gt, "Heci Packet",
2 * PXP43_MAX_HECI_INOUT_SIZE,
&exec_res->pkt_vma, &exec_res->pkt_vaddr);
if (err)
return err;
err = gsccs_create_buffer(pxp->ctrl_gt, "Batch Buffer", PAGE_SIZE,
&exec_res->bb_vma, &exec_res->bb_vaddr);
if (err)
goto free_pkt;
/* Finally, create an intel_context to be used during the submission */
ce = intel_context_create(engine);
if (IS_ERR(ce)) {
drm_err(&gt->i915->drm, "Failed creating gsccs backend ctx\n");
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
err = PTR_ERR(ce);
goto free_batch;
}
i915_vm_put(ce->vm);
ce->vm = i915_vm_get(pxp->ctrl_gt->vm);
exec_res->ce = ce;
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
/* initialize host-session-handle (for all i915-to-gsc-firmware PXP cmds) */
get_random_bytes(&exec_res->host_session_handle, sizeof(exec_res->host_session_handle));
return 0;
drm/i915/pxp: Add GSC-CS backend to send GSC fw messages Add GSC engine based method for sending PXP firmware packets to the GSC firmware for MTL (and future) products. Use the newly added helpers to populate the GSC-CS memory header and send the message packet to the FW by dispatching the GSC_HECI_CMD_PKT instruction on the GSC engine. We use non-priveleged batches for submission to GSC engine which require two buffers for the request: - a buffer for the HECI packet that contains PXP FW commands - a batch-buffer that contains the engine instruction for sending the HECI packet to the GSC firmware. Thus, add the allocation and freeing of these buffers in gsccs init and fini. The GSC-fw may reply to commands with a SUCCESS but with an additional pending-bit set in the reply packet. This bit means the GSC-FW is currently busy and the caller needs to try again with the gsc_message_handle the fw returned. Thus, add a wrapper to continuously retry send_message while replaying the gsc_message_handle. Retries need to follow the arch-spec count and delay until GSC-FW replies with the real SUCCESS or timeout after that spec'd delay. The GSC-fw requires a non-zero host_session_handle provided by the caller to enable gsc_message_handle tracking. Thus, allocate the host_session_handle at init and destroy it at fini (the latter requiring an FYI to the gsc-firmware). Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Signed-off-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230511231738.1077674-5-alan.previn.teres.alexis@intel.com
2023-05-11 16:17:34 -07:00
free_batch:
i915_vma_unpin_and_release(&exec_res->bb_vma, I915_VMA_RELEASE_MAP);
free_pkt:
i915_vma_unpin_and_release(&exec_res->pkt_vma, I915_VMA_RELEASE_MAP);
memset(exec_res, 0, sizeof(*exec_res));
return err;
}
void intel_pxp_gsccs_fini(struct intel_pxp *pxp)
{
intel_wakeref_t wakeref;
gsccs_destroy_execution_resource(pxp);
with_intel_runtime_pm(&pxp->ctrl_gt->i915->runtime_pm, wakeref)
intel_pxp_fini_hw(pxp);
}
int intel_pxp_gsccs_init(struct intel_pxp *pxp)
{
int ret;
intel_wakeref_t wakeref;
ret = gsccs_allocate_execution_resource(pxp);
if (!ret) {
with_intel_runtime_pm(&pxp->ctrl_gt->i915->runtime_pm, wakeref)
intel_pxp_init_hw(pxp);
}
return ret;
}