github-mirror/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-05-29 17:43:52 +02:00
Code Issues Actions Packages Projects Releases Wiki Activity Graph

UAPI Changes:

Browse Source 
Limit number of jobs per exec queue (Shuicheng)
 Add sriov_admin sysfs tree (Michal)
 
 Driver Changes:
 
 Fix an uninitialized value (Thomas)
 Expose a residency counter through debugfs (Mohammed Thasleem)
 Workaround enabling and improvement (Tapani, Tangudu)
 More Crescent Island-specific support (Sk Anirban, Lucas)
 PAT entry dump improvement (Xin)
 Inline gt_reset in the worker (Lucas)
 Synchronize GT reset with device unbind (Balasubramani)
 Do clean shutdown also when using flr (Jouni)
 Fix serialization on burst of unbinds (Matt Brost)
 Pagefault Refactor (Matt Brost)
 Remove some unused code (Gwan-gyeong)
 Avoid TOCTOU when montoring throttle reasons (Lucas)
 Add/extend workaround (Nitin)
 SRIOV migration work / plumbing (Michal Wajdeczko, Michal Winiarski, Lukasz)
 Drop debug flag requirement for VF resource fixup
 Fix MTL vm_max_level (Rodrigo)
 Changes around TILE_ADDR_RANGE for platform compatibility
 (Fei, Lucas)
 Add runtime registers for GFX ver >= 35 (Piotr)
 Kerneldoc fix (Kriish)
 Rework pcode error mapping (Lucas)
 Allow lockdown the PF (Michal)
 Eliminate GUC code caching of some frequency values (Sk)
 Improvements around forcewake referencing (Matt Roper)
 -----BEGIN PGP SIGNATURE-----
 
 iHUEABYKAB0WIQRskUM7w1oG5rx2IZO4FpNVCsYGvwUCaRcIGwAKCRC4FpNVCsYG
 v3VWAQD7eVdaBKgOUoyWexzeWiHeAEAe9QwgTdK71PBNHF1kIQEA3dg12QDETxFp
 pWvOYie6bl+ApXx/JJ0OKtMeIFz2mQc=
 =MS1G
 -----END PGP SIGNATURE-----

Merge tag 'drm-xe-next-2025-11-14' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next

Driver Changes:

Avoid TOCTOU when montoring throttle reasons (Lucas)
Add/extend workaround (Nitin)
SRIOV migration work / plumbing (Michal Wajdeczko, Michal Winiarski, Lukasz)
Drop debug flag requirement for VF resource fixup
Fix MTL vm_max_level (Rodrigo)
Changes around TILE_ADDR_RANGE for platform compatibility
(Fei, Lucas)
Add runtime registers for GFX ver >= 35 (Piotr)
Kerneldoc fix (Kriish)
Rework pcode error mapping (Lucas)
Allow lockdown the PF (Michal)
Eliminate GUC code caching of some frequency values (Sk)
Improvements around forcewake referencing (Matt Roper)

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Link: https://patch.msgid.link/aRcJOrisG2qPbucE@fedora
This commit is contained in:
Dave Airlie 2025-11-17 13:39:45 +10:00
commit 727bf2dc96
55 changed files with 4038 additions and 490 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
drivers/gpu/drm/xe/Makefile
View File

@ -174,9 +174,11 @@ xe-$(CONFIG_PCI_IOV) += \
xe_lmtt_2l.o \
xe_lmtt_ml.o \
xe_pci_sriov.o \
xe_sriov_packet.o \
xe_sriov_pf.o \
xe_sriov_pf_control.o \
xe_sriov_pf_debugfs.o \
xe_sriov_pf_migration.o \
xe_sriov_pf_provision.o \
xe_sriov_pf_service.o \
xe_sriov_pf_sysfs.o \

1
drivers/gpu/drm/xe/regs/xe_gt_regs.h
View File

@ -101,7 +101,6 @@
#define XE2_LMEM_CFG XE_REG(0x48b0)
#define XEHP_TILE_ADDR_RANGE(_idx) XE_REG_MCR(0x4900 + (_idx) * 4)
#define XEHP_FLAT_CCS_BASE_ADDR XE_REG_MCR(0x4910)
#define XEHP_FLAT_CCS_PTR REG_GENMASK(31, 8)

2
drivers/gpu/drm/xe/regs/xe_regs.h
View File

@ -40,6 +40,8 @@
#define STOLEN_RESERVED XE_REG(0x1082c0)
#define WOPCM_SIZE_MASK REG_GENMASK64(9, 7)
#define SG_TILE_ADDR_RANGE(_idx) XE_REG(0x1083a0 + (_idx) * 4)
#define MTL_RP_STATE_CAP XE_REG(0x138000)
#define MTL_GT_RPA_FREQUENCY XE_REG(0x138008)

208
drivers/gpu/drm/xe/tests/xe_gt_sriov_pf_config_kunit.c Normal file
View File

@ -0,0 +1,208 @@
// SPDX-License-Identifier: GPL-2.0 AND MIT
/*
* Copyright © 2025 Intel Corporation
*/
#include <kunit/static_stub.h>
#include <kunit/test.h>
#include <kunit/test-bug.h>
#include "xe_kunit_helpers.h"
#include "xe_pci_test.h"
#define TEST_MAX_VFS 63
static void pf_set_admin_mode(struct xe_device *xe, bool enable)
{
/* should match logic of xe_sriov_pf_admin_only() */
xe->info.probe_display = !enable;
KUNIT_EXPECT_EQ(kunit_get_current_test(), enable, xe_sriov_pf_admin_only(xe));
}
static const void *num_vfs_gen_param(struct kunit *test, const void *prev, char *desc)
{
unsigned long next = 1 + (unsigned long)prev;
if (next > TEST_MAX_VFS)
return NULL;
snprintf(desc, KUNIT_PARAM_DESC_SIZE, "%lu VF%s",
next, str_plural(next));
return (void *)next;
}
static int pf_gt_config_test_init(struct kunit *test)
{
struct xe_pci_fake_data fake = {
.sriov_mode = XE_SRIOV_MODE_PF,
.platform = XE_TIGERLAKE, /* any random platform with SR-IOV */
.subplatform = XE_SUBPLATFORM_NONE,
};
struct xe_device *xe;
struct xe_gt *gt;
test->priv = &fake;
xe_kunit_helper_xe_device_test_init(test);
xe = test->priv;
KUNIT_ASSERT_TRUE(test, IS_SRIOV_PF(xe));
gt = xe_root_mmio_gt(xe);
KUNIT_ASSERT_NOT_ERR_OR_NULL(test, gt);
test->priv = gt;
/* pretend it can support up to 63 VFs */
xe->sriov.pf.device_total_vfs = TEST_MAX_VFS;
xe->sriov.pf.driver_max_vfs = TEST_MAX_VFS;
KUNIT_ASSERT_EQ(test, xe_sriov_pf_get_totalvfs(xe), 63);
pf_set_admin_mode(xe, false);
KUNIT_ASSERT_EQ(test, xe_sriov_init(xe), 0);
/* more sanity checks */
KUNIT_EXPECT_EQ(test, GUC_ID_MAX + 1, SZ_64K);
KUNIT_EXPECT_EQ(test, GUC_NUM_DOORBELLS, SZ_256);
return 0;
}
static void fair_contexts_1vf(struct kunit *test)
{
struct xe_gt *gt = test->priv;
struct xe_device *xe = gt_to_xe(gt);
pf_set_admin_mode(xe, false);
KUNIT_ASSERT_FALSE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_EQ(test, SZ_32K, pf_profile_fair_ctxs(gt, 1));
pf_set_admin_mode(xe, true);
KUNIT_ASSERT_TRUE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_EQ(test, SZ_64K - SZ_1K, pf_profile_fair_ctxs(gt, 1));
}
static void fair_contexts(struct kunit *test)
{
unsigned int num_vfs = (unsigned long)test->param_value;
struct xe_gt *gt = test->priv;
struct xe_device *xe = gt_to_xe(gt);
pf_set_admin_mode(xe, false);
KUNIT_ASSERT_FALSE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_TRUE(test, is_power_of_2(pf_profile_fair_ctxs(gt, num_vfs)));
KUNIT_EXPECT_GT(test, GUC_ID_MAX, num_vfs * pf_profile_fair_ctxs(gt, num_vfs));
if (num_vfs > 31)
KUNIT_ASSERT_EQ(test, SZ_1K, pf_profile_fair_ctxs(gt, num_vfs));
else if (num_vfs > 15)
KUNIT_ASSERT_EQ(test, SZ_2K, pf_profile_fair_ctxs(gt, num_vfs));
else if (num_vfs > 7)
KUNIT_ASSERT_EQ(test, SZ_4K, pf_profile_fair_ctxs(gt, num_vfs));
else if (num_vfs > 3)
KUNIT_ASSERT_EQ(test, SZ_8K, pf_profile_fair_ctxs(gt, num_vfs));
else if (num_vfs > 1)
KUNIT_ASSERT_EQ(test, SZ_16K, pf_profile_fair_ctxs(gt, num_vfs));
else
KUNIT_ASSERT_EQ(test, SZ_32K, pf_profile_fair_ctxs(gt, num_vfs));
}
static void fair_doorbells_1vf(struct kunit *test)
{
struct xe_gt *gt = test->priv;
struct xe_device *xe = gt_to_xe(gt);
pf_set_admin_mode(xe, false);
KUNIT_ASSERT_FALSE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_EQ(test, 128, pf_profile_fair_dbs(gt, 1));
pf_set_admin_mode(xe, true);
KUNIT_ASSERT_TRUE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_EQ(test, 240, pf_profile_fair_dbs(gt, 1));
}
static void fair_doorbells(struct kunit *test)
{
unsigned int num_vfs = (unsigned long)test->param_value;
struct xe_gt *gt = test->priv;
struct xe_device *xe = gt_to_xe(gt);
pf_set_admin_mode(xe, false);
KUNIT_ASSERT_FALSE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_TRUE(test, is_power_of_2(pf_profile_fair_dbs(gt, num_vfs)));
KUNIT_EXPECT_GE(test, GUC_NUM_DOORBELLS, (num_vfs + 1) * pf_profile_fair_dbs(gt, num_vfs));
if (num_vfs > 31)
KUNIT_ASSERT_EQ(test, SZ_4, pf_profile_fair_dbs(gt, num_vfs));
else if (num_vfs > 15)
KUNIT_ASSERT_EQ(test, SZ_8, pf_profile_fair_dbs(gt, num_vfs));
else if (num_vfs > 7)
KUNIT_ASSERT_EQ(test, SZ_16, pf_profile_fair_dbs(gt, num_vfs));
else if (num_vfs > 3)
KUNIT_ASSERT_EQ(test, SZ_32, pf_profile_fair_dbs(gt, num_vfs));
else if (num_vfs > 1)
KUNIT_ASSERT_EQ(test, SZ_64, pf_profile_fair_dbs(gt, num_vfs));
else
KUNIT_ASSERT_EQ(test, SZ_128, pf_profile_fair_dbs(gt, num_vfs));
}
static void fair_ggtt_1vf(struct kunit *test)
{
struct xe_gt *gt = test->priv;
struct xe_device *xe = gt_to_xe(gt);
pf_set_admin_mode(xe, false);
KUNIT_ASSERT_FALSE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_EQ(test, SZ_2G, pf_profile_fair_ggtt(gt, 1));
pf_set_admin_mode(xe, true);
KUNIT_ASSERT_TRUE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_EQ(test, SZ_2G + SZ_1G + SZ_512M, pf_profile_fair_ggtt(gt, 1));
}
static void fair_ggtt(struct kunit *test)
{
unsigned int num_vfs = (unsigned long)test->param_value;
struct xe_gt *gt = test->priv;
struct xe_device *xe = gt_to_xe(gt);
u64 alignment = pf_get_ggtt_alignment(gt);
u64 shareable = SZ_2G + SZ_1G + SZ_512M;
pf_set_admin_mode(xe, false);
KUNIT_ASSERT_FALSE(test, xe_sriov_pf_admin_only(xe));
KUNIT_EXPECT_TRUE(test, IS_ALIGNED(pf_profile_fair_ggtt(gt, num_vfs), alignment));
KUNIT_EXPECT_GE(test, shareable, num_vfs * pf_profile_fair_ggtt(gt, num_vfs));
if (num_vfs > 56)
KUNIT_ASSERT_EQ(test, SZ_64M - SZ_8M, pf_profile_fair_ggtt(gt, num_vfs));
else if (num_vfs > 28)
KUNIT_ASSERT_EQ(test, SZ_64M, pf_profile_fair_ggtt(gt, num_vfs));
else if (num_vfs > 14)
KUNIT_ASSERT_EQ(test, SZ_128M, pf_profile_fair_ggtt(gt, num_vfs));
else if (num_vfs > 7)
KUNIT_ASSERT_EQ(test, SZ_256M, pf_profile_fair_ggtt(gt, num_vfs));
else if (num_vfs > 3)
KUNIT_ASSERT_EQ(test, SZ_512M, pf_profile_fair_ggtt(gt, num_vfs));
else if (num_vfs > 1)
KUNIT_ASSERT_EQ(test, SZ_1G, pf_profile_fair_ggtt(gt, num_vfs));
else
KUNIT_ASSERT_EQ(test, SZ_2G, pf_profile_fair_ggtt(gt, num_vfs));
}
static struct kunit_case pf_gt_config_test_cases[] = {
KUNIT_CASE(fair_contexts_1vf),
KUNIT_CASE(fair_doorbells_1vf),
KUNIT_CASE(fair_ggtt_1vf),
KUNIT_CASE_PARAM(fair_contexts, num_vfs_gen_param),
KUNIT_CASE_PARAM(fair_doorbells, num_vfs_gen_param),
KUNIT_CASE_PARAM(fair_ggtt, num_vfs_gen_param),
{}
};
static struct kunit_suite pf_gt_config_suite = {
.name = "pf_gt_config",
.test_cases = pf_gt_config_test_cases,
.init = pf_gt_config_test_init,
};
kunit_test_suite(pf_gt_config_suite);

8
drivers/gpu/drm/xe/xe_eu_stall.c
View File

@ -49,6 +49,7 @@ struct xe_eu_stall_data_stream {
wait_queue_head_t poll_wq;
size_t data_record_size;
size_t per_xecore_buf_size;
unsigned int fw_ref;
struct xe_gt *gt;
struct xe_bo *bo;
@ -660,13 +661,12 @@ static int xe_eu_stall_stream_enable(struct xe_eu_stall_data_stream *stream)
struct per_xecore_buf *xecore_buf;
struct xe_gt *gt = stream->gt;
u16 group, instance;
unsigned int fw_ref;
int xecore;
/* Take runtime pm ref and forcewake to disable RC6 */
xe_pm_runtime_get(gt_to_xe(gt));
fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FW_RENDER);
if (!xe_force_wake_ref_has_domain(fw_ref, XE_FW_RENDER)) {
stream->fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FW_RENDER);
if (!xe_force_wake_ref_has_domain(stream->fw_ref, XE_FW_RENDER)) {
xe_gt_err(gt, "Failed to get RENDER forcewake\n");
xe_pm_runtime_put(gt_to_xe(gt));
return -ETIMEDOUT;
@ -832,7 +832,7 @@ static int xe_eu_stall_disable_locked(struct xe_eu_stall_data_stream *stream)
xe_gt_mcr_multicast_write(gt, ROW_CHICKEN2,
_MASKED_BIT_DISABLE(DISABLE_DOP_GATING));
xe_force_wake_put(gt_to_fw(gt), XE_FW_RENDER);
xe_force_wake_put(gt_to_fw(gt), stream->fw_ref);
xe_pm_runtime_put(gt_to_xe(gt));
return 0;

26
drivers/gpu/drm/xe/xe_force_wake_types.h
View File

@ -52,7 +52,22 @@ enum xe_force_wake_domains {
};
/**
* struct xe_force_wake_domain - Xe force wake domains
* struct xe_force_wake_domain - Xe force wake power domain
*
* Represents an individual device-internal power domain. The driver must
* ensure the power domain is awake before accessing registers or other
* hardware functionality that is part of the power domain. Since different
* driver threads may access hardware units simultaneously, a reference count
* is used to ensure that the domain remains awake as long as any software
* is using the part of the hardware covered by the power domain.
*
* Hardware provides a register interface to allow the driver to request
* wake/sleep of power domains, although in most cases the actual action of
* powering the hardware up/down is handled by firmware (and may be subject to
* requirements and constraints outside of the driver's visibility) so the
* driver needs to wait for an acknowledgment that a wake request has been
* acted upon before accessing the parts of the hardware that reside within the
* power domain.
*/
struct xe_force_wake_domain {
/** @id: domain force wake id */
@ -70,7 +85,14 @@ struct xe_force_wake_domain {
};
/**
* struct xe_force_wake - Xe force wake
* struct xe_force_wake - Xe force wake collection
*
* Represents a collection of related power domains (struct
* xe_force_wake_domain) associated with a subunit of the device.
*
* Currently only used for GT power domains (where the term "forcewake" is used
* in the hardware documentation), although the interface could be extended to
* power wells in other parts of the hardware in the future.
*/
struct xe_force_wake {
/** @gt: back pointers to GT */

104
drivers/gpu/drm/xe/xe_ggtt.c
View File

@ -151,6 +151,14 @@ static void xe_ggtt_set_pte_and_flush(struct xe_ggtt *ggtt, u64 addr, u64 pte)
ggtt_update_access_counter(ggtt);
}
static u64 xe_ggtt_get_pte(struct xe_ggtt *ggtt, u64 addr)
{
xe_tile_assert(ggtt->tile, !(addr & XE_PTE_MASK));
xe_tile_assert(ggtt->tile, addr < ggtt->size);
return readq(&ggtt->gsm[addr >> XE_PTE_SHIFT]);
}
static void xe_ggtt_clear(struct xe_ggtt *ggtt, u64 start, u64 size)
{
u16 pat_index = tile_to_xe(ggtt->tile)->pat.idx[XE_CACHE_WB];
@ -233,16 +241,19 @@ void xe_ggtt_might_lock(struct xe_ggtt *ggtt)
static const struct xe_ggtt_pt_ops xelp_pt_ops = {
.pte_encode_flags = xelp_ggtt_pte_flags,
.ggtt_set_pte = xe_ggtt_set_pte,
.ggtt_get_pte = xe_ggtt_get_pte,
};
static const struct xe_ggtt_pt_ops xelpg_pt_ops = {
.pte_encode_flags = xelpg_ggtt_pte_flags,
.ggtt_set_pte = xe_ggtt_set_pte,
.ggtt_get_pte = xe_ggtt_get_pte,
};
static const struct xe_ggtt_pt_ops xelpg_pt_wa_ops = {
.pte_encode_flags = xelpg_ggtt_pte_flags,
.ggtt_set_pte = xe_ggtt_set_pte_and_flush,
.ggtt_get_pte = xe_ggtt_get_pte,
};
static void __xe_ggtt_init_early(struct xe_ggtt *ggtt, u32 reserved)
@ -697,6 +708,20 @@ bool xe_ggtt_node_allocated(const struct xe_ggtt_node *node)
return drm_mm_node_allocated(&node->base);
}
/**
* xe_ggtt_node_pt_size() - Get the size of page table entries needed to map a GGTT node.
* @node: the &xe_ggtt_node
*
* Return: GGTT node page table entries size in bytes.
*/
size_t xe_ggtt_node_pt_size(const struct xe_ggtt_node *node)
{
if (!node)
return 0;
return node->base.size / XE_PAGE_SIZE * sizeof(u64);
}
/**
* xe_ggtt_map_bo - Map the BO into GGTT
* @ggtt: the &xe_ggtt where node will be mapped
@ -930,6 +955,85 @@ void xe_ggtt_assign(const struct xe_ggtt_node *node, u16 vfid)
xe_ggtt_assign_locked(node->ggtt, &node->base, vfid);
mutex_unlock(&node->ggtt->lock);
}
/**
* xe_ggtt_node_save() - Save a &xe_ggtt_node to a buffer.
* @node: the &xe_ggtt_node to be saved
* @dst: destination buffer
* @size: destination buffer size in bytes
* @vfid: VF identifier
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_ggtt_node_save(struct xe_ggtt_node *node, void *dst, size_t size, u16 vfid)
{
struct xe_ggtt *ggtt;
u64 start, end;
u64 *buf = dst;
u64 pte;
if (!node)
return -ENOENT;
guard(mutex)(&node->ggtt->lock);
if (xe_ggtt_node_pt_size(node) != size)
return -EINVAL;
ggtt = node->ggtt;
start = node->base.start;
end = start + node->base.size - 1;
while (start < end) {
pte = ggtt->pt_ops->ggtt_get_pte(ggtt, start);
if (vfid != u64_get_bits(pte, GGTT_PTE_VFID))
return -EPERM;
*buf++ = u64_replace_bits(pte, 0, GGTT_PTE_VFID);
start += XE_PAGE_SIZE;
}
return 0;
}
/**
* xe_ggtt_node_load() - Load a &xe_ggtt_node from a buffer.
* @node: the &xe_ggtt_node to be loaded
* @src: source buffer
* @size: source buffer size in bytes
* @vfid: VF identifier
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_ggtt_node_load(struct xe_ggtt_node *node, const void *src, size_t size, u16 vfid)
{
u64 vfid_pte = xe_encode_vfid_pte(vfid);
const u64 *buf = src;
struct xe_ggtt *ggtt;
u64 start, end;
if (!node)
return -ENOENT;
guard(mutex)(&node->ggtt->lock);
if (xe_ggtt_node_pt_size(node) != size)
return -EINVAL;
ggtt = node->ggtt;
start = node->base.start;
end = start + node->base.size - 1;
while (start < end) {
vfid_pte = u64_replace_bits(*buf++, vfid, GGTT_PTE_VFID);
ggtt->pt_ops->ggtt_set_pte(ggtt, start, vfid_pte);
start += XE_PAGE_SIZE;
}
xe_ggtt_invalidate(ggtt);
return 0;
}
#endif
/**

3
drivers/gpu/drm/xe/xe_ggtt.h
View File

@ -29,6 +29,7 @@ int xe_ggtt_node_insert_locked(struct xe_ggtt_node *node,
u32 size, u32 align, u32 mm_flags);
void xe_ggtt_node_remove(struct xe_ggtt_node *node, bool invalidate);
bool xe_ggtt_node_allocated(const struct xe_ggtt_node *node);
size_t xe_ggtt_node_pt_size(const struct xe_ggtt_node *node);
void xe_ggtt_map_bo(struct xe_ggtt *ggtt, struct xe_ggtt_node *node,
struct xe_bo *bo, u16 pat_index);
void xe_ggtt_map_bo_unlocked(struct xe_ggtt *ggtt, struct xe_bo *bo);
@ -43,6 +44,8 @@ u64 xe_ggtt_print_holes(struct xe_ggtt *ggtt, u64 alignment, struct drm_printer
#ifdef CONFIG_PCI_IOV
void xe_ggtt_assign(const struct xe_ggtt_node *node, u16 vfid);
int xe_ggtt_node_save(struct xe_ggtt_node *node, void *dst, size_t size, u16 vfid);
int xe_ggtt_node_load(struct xe_ggtt_node *node, const void *src, size_t size, u16 vfid);
#endif
#ifndef CONFIG_LOCKDEP

2
drivers/gpu/drm/xe/xe_ggtt_types.h
View File

@ -78,6 +78,8 @@ struct xe_ggtt_pt_ops {
u64 (*pte_encode_flags)(struct xe_bo *bo, u16 pat_index);
/** @ggtt_set_pte: Directly write into GGTT's PTE */
void (*ggtt_set_pte)(struct xe_ggtt *ggtt, u64 addr, u64 pte);
/** @ggtt_get_pte: Directly read from GGTT's PTE */
u64 (*ggtt_get_pte)(struct xe_ggtt *ggtt, u64 addr);
};
#endif

143
drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c
View File

@ -9,6 +9,7 @@
#include "abi/guc_actions_sriov_abi.h"
#include "abi/guc_klvs_abi.h"
#include "regs/xe_gtt_defs.h"
#include "regs/xe_guc_regs.h"
#include "xe_bo.h"
@ -697,6 +698,22 @@ static u64 pf_estimate_fair_ggtt(struct xe_gt *gt, unsigned int num_vfs)
return fair;
}
static u64 pf_profile_fair_ggtt(struct xe_gt *gt, unsigned int num_vfs)
{
bool admin_only_pf = xe_sriov_pf_admin_only(gt_to_xe(gt));
u64 shareable = ALIGN_DOWN(GUC_GGTT_TOP, SZ_512M);
u64 alignment = pf_get_ggtt_alignment(gt);
if (admin_only_pf && num_vfs == 1)
return ALIGN_DOWN(shareable, alignment);
/* need to hardcode due to ~512M of GGTT being reserved */
if (num_vfs > 56)
return SZ_64M - SZ_8M;
return rounddown_pow_of_two(shareable / num_vfs);
}
/**
* xe_gt_sriov_pf_config_set_fair_ggtt - Provision many VFs with fair GGTT.
* @gt: the &xe_gt (can't be media)
@ -710,6 +727,7 @@ static u64 pf_estimate_fair_ggtt(struct xe_gt *gt, unsigned int num_vfs)
int xe_gt_sriov_pf_config_set_fair_ggtt(struct xe_gt *gt, unsigned int vfid,
unsigned int num_vfs)
{
u64 profile = pf_profile_fair_ggtt(gt, num_vfs);
u64 fair;
xe_gt_assert(gt, vfid);
@ -723,9 +741,71 @@ int xe_gt_sriov_pf_config_set_fair_ggtt(struct xe_gt *gt, unsigned int vfid,
if (!fair)
return -ENOSPC;
fair = min(fair, profile);
if (fair < profile)
xe_gt_sriov_info(gt, "Using non-profile provisioning (%s %llu vs %llu)\n",
"GGTT", fair, profile);
return xe_gt_sriov_pf_config_bulk_set_ggtt(gt, vfid, num_vfs, fair);
}
/**
* xe_gt_sriov_pf_config_ggtt_save() - Save a VF provisioned GGTT data into a buffer.
* @gt: the &xe_gt
* @vfid: VF identifier (can't be 0)
* @buf: the GGTT data destination buffer (or NULL to query the buf size)
* @size: the size of the buffer (or 0 to query the buf size)
*
* This function can only be called on PF.
*
* Return: size of the buffer needed to save GGTT data if querying,
* 0 on successful save or a negative error code on failure.
*/
ssize_t xe_gt_sriov_pf_config_ggtt_save(struct xe_gt *gt, unsigned int vfid,
void *buf, size_t size)
{
struct xe_ggtt_node *node;
xe_gt_assert(gt, IS_SRIOV_PF(gt_to_xe(gt)));
xe_gt_assert(gt, vfid);
xe_gt_assert(gt, !(!buf ^ !size));
guard(mutex)(xe_gt_sriov_pf_master_mutex(gt));
node = pf_pick_vf_config(gt, vfid)->ggtt_region;
if (!buf)
return xe_ggtt_node_pt_size(node);
return xe_ggtt_node_save(node, buf, size, vfid);
}
/**
* xe_gt_sriov_pf_config_ggtt_restore() - Restore a VF provisioned GGTT data from a buffer.
* @gt: the &xe_gt
* @vfid: VF identifier (can't be 0)
* @buf: the GGTT data source buffer
* @size: the size of the buffer
*
* This function can only be called on PF.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_config_ggtt_restore(struct xe_gt *gt, unsigned int vfid,
const void *buf, size_t size)
{
struct xe_ggtt_node *node;
xe_gt_assert(gt, IS_SRIOV_PF(gt_to_xe(gt)));
xe_gt_assert(gt, vfid);
guard(mutex)(xe_gt_sriov_pf_master_mutex(gt));
node = pf_pick_vf_config(gt, vfid)->ggtt_region;
return xe_ggtt_node_load(node, buf, size, vfid);
}
static u32 pf_get_min_spare_ctxs(struct xe_gt *gt)
{
/* XXX: preliminary */
@ -985,6 +1065,16 @@ int xe_gt_sriov_pf_config_bulk_set_ctxs(struct xe_gt *gt, unsigned int vfid,
"GuC context IDs", no_unit, n, err);
}
static u32 pf_profile_fair_ctxs(struct xe_gt *gt, unsigned int num_vfs)
{
bool admin_only_pf = xe_sriov_pf_admin_only(gt_to_xe(gt));
if (admin_only_pf && num_vfs == 1)
return ALIGN_DOWN(GUC_ID_MAX, SZ_1K);
return rounddown_pow_of_two(GUC_ID_MAX / num_vfs);
}
static u32 pf_estimate_fair_ctxs(struct xe_gt *gt, unsigned int num_vfs)
{
struct xe_guc_id_mgr *idm = &gt->uc.guc.submission_state.idm;
@ -1017,6 +1107,7 @@ static u32 pf_estimate_fair_ctxs(struct xe_gt *gt, unsigned int num_vfs)
int xe_gt_sriov_pf_config_set_fair_ctxs(struct xe_gt *gt, unsigned int vfid,
unsigned int num_vfs)
{
u32 profile = pf_profile_fair_ctxs(gt, num_vfs);
u32 fair;
xe_gt_assert(gt, vfid);
@ -1029,6 +1120,11 @@ int xe_gt_sriov_pf_config_set_fair_ctxs(struct xe_gt *gt, unsigned int vfid,
if (!fair)
return -ENOSPC;
fair = min(fair, profile);
if (fair < profile)
xe_gt_sriov_info(gt, "Using non-profile provisioning (%s %u vs %u)\n",
"GuC context IDs", fair, profile);
return xe_gt_sriov_pf_config_bulk_set_ctxs(gt, vfid, num_vfs, fair);
}
@ -1233,6 +1329,17 @@ int xe_gt_sriov_pf_config_bulk_set_dbs(struct xe_gt *gt, unsigned int vfid,
"GuC doorbell IDs", no_unit, n, err);
}
static u32 pf_profile_fair_dbs(struct xe_gt *gt, unsigned int num_vfs)
{
bool admin_only_pf = xe_sriov_pf_admin_only(gt_to_xe(gt));
/* XXX: preliminary */
if (admin_only_pf && num_vfs == 1)
return GUC_NUM_DOORBELLS - SZ_16;
return rounddown_pow_of_two(GUC_NUM_DOORBELLS / (num_vfs + 1));
}
static u32 pf_estimate_fair_dbs(struct xe_gt *gt, unsigned int num_vfs)
{
struct xe_guc_db_mgr *dbm = &gt->uc.guc.dbm;
@ -1265,6 +1372,7 @@ static u32 pf_estimate_fair_dbs(struct xe_gt *gt, unsigned int num_vfs)
int xe_gt_sriov_pf_config_set_fair_dbs(struct xe_gt *gt, unsigned int vfid,
unsigned int num_vfs)
{
u32 profile = pf_profile_fair_dbs(gt, num_vfs);
u32 fair;
xe_gt_assert(gt, vfid);
@ -1277,6 +1385,11 @@ int xe_gt_sriov_pf_config_set_fair_dbs(struct xe_gt *gt, unsigned int vfid,
if (!fair)
return -ENOSPC;
fair = min(fair, profile);
if (fair < profile)
xe_gt_sriov_info(gt, "Using non-profile provisioning (%s %u vs %u)\n",
"GuC doorbell IDs", fair, profile);
return xe_gt_sriov_pf_config_bulk_set_dbs(gt, vfid, num_vfs, fair);
}
@ -1602,6 +1715,32 @@ int xe_gt_sriov_pf_config_bulk_set_lmem(struct xe_gt *gt, unsigned int vfid,
"LMEM", n, err);
}
static struct xe_bo *pf_get_vf_config_lmem_obj(struct xe_gt *gt, unsigned int vfid)
{
struct xe_gt_sriov_config *config = pf_pick_vf_config(gt, vfid);
return config->lmem_obj;
}
/**
* xe_gt_sriov_pf_config_get_lmem_obj() - Take a reference to the struct &xe_bo backing VF LMEM.
* @gt: the &xe_gt
* @vfid: the VF identifier (can't be 0)
*
* This function can only be called on PF.
* The caller is responsible for calling xe_bo_put() on the returned object.
*
* Return: pointer to struct &xe_bo backing VF LMEM (if any).
*/
struct xe_bo *xe_gt_sriov_pf_config_get_lmem_obj(struct xe_gt *gt, unsigned int vfid)
{
xe_gt_assert(gt, vfid);
guard(mutex)(xe_gt_sriov_pf_master_mutex(gt));
return xe_bo_get(pf_get_vf_config_lmem_obj(gt, vfid));
}
static u64 pf_query_free_lmem(struct xe_gt *gt)
{
struct xe_tile *tile = gt->tile;
@ -2793,3 +2932,7 @@ int xe_gt_sriov_pf_config_print_available_ggtt(struct xe_gt *gt, struct drm_prin
return 0;
}
#if IS_BUILTIN(CONFIG_DRM_XE_KUNIT_TEST)
#include "tests/xe_gt_sriov_pf_config_kunit.c"
#endif

6
drivers/gpu/drm/xe/xe_gt_sriov_pf_config.h
View File

@ -36,6 +36,7 @@ int xe_gt_sriov_pf_config_set_lmem(struct xe_gt *gt, unsigned int vfid, u64 size
int xe_gt_sriov_pf_config_set_fair_lmem(struct xe_gt *gt, unsigned int vfid, unsigned int num_vfs);
int xe_gt_sriov_pf_config_bulk_set_lmem(struct xe_gt *gt, unsigned int vfid, unsigned int num_vfs,
u64 size);
struct xe_bo *xe_gt_sriov_pf_config_get_lmem_obj(struct xe_gt *gt, unsigned int vfid);
u32 xe_gt_sriov_pf_config_get_exec_quantum(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_config_set_exec_quantum(struct xe_gt *gt, unsigned int vfid, u32 exec_quantum);
@ -71,6 +72,11 @@ ssize_t xe_gt_sriov_pf_config_save(struct xe_gt *gt, unsigned int vfid, void *bu
int xe_gt_sriov_pf_config_restore(struct xe_gt *gt, unsigned int vfid,
const void *buf, size_t size);
ssize_t xe_gt_sriov_pf_config_ggtt_save(struct xe_gt *gt, unsigned int vfid,
void *buf, size_t size);
int xe_gt_sriov_pf_config_ggtt_restore(struct xe_gt *gt, unsigned int vfid,
const void *buf, size_t size);
bool xe_gt_sriov_pf_config_is_empty(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_config_init(struct xe_gt *gt);

650
drivers/gpu/drm/xe/xe_gt_sriov_pf_control.c
View File

@ -18,7 +18,10 @@
#include "xe_gt_sriov_printk.h"
#include "xe_guc_ct.h"
#include "xe_sriov.h"
#include "xe_sriov_packet.h"
#include "xe_sriov_packet_types.h"
#include "xe_sriov_pf_control.h"
#include "xe_sriov_pf_migration.h"
#include "xe_sriov_pf_service.h"
#include "xe_tile.h"
@ -181,9 +184,20 @@ static const char *control_bit_to_string(enum xe_gt_sriov_control_bits bit)
CASE2STR(PAUSE_SEND_PAUSE);
CASE2STR(PAUSE_WAIT_GUC);
CASE2STR(PAUSE_GUC_DONE);
CASE2STR(PAUSE_SAVE_GUC);
CASE2STR(PAUSE_FAILED);
CASE2STR(PAUSED);
CASE2STR(SAVE_WIP);
CASE2STR(SAVE_PROCESS_DATA);
CASE2STR(SAVE_WAIT_DATA);
CASE2STR(SAVE_DATA_DONE);
CASE2STR(SAVE_FAILED);
CASE2STR(SAVED);
CASE2STR(RESTORE_WIP);
CASE2STR(RESTORE_PROCESS_DATA);
CASE2STR(RESTORE_WAIT_DATA);
CASE2STR(RESTORE_DATA_DONE);
CASE2STR(RESTORE_FAILED);
CASE2STR(RESTORED);
CASE2STR(RESUME_WIP);
CASE2STR(RESUME_SEND_RESUME);
CASE2STR(RESUME_FAILED);
@ -208,6 +222,8 @@ static unsigned long pf_get_default_timeout(enum xe_gt_sriov_control_bits bit)
case XE_GT_SRIOV_STATE_FLR_WIP:
case XE_GT_SRIOV_STATE_FLR_RESET_CONFIG:
return 5 * HZ;
case XE_GT_SRIOV_STATE_RESTORE_WIP:
return 20 * HZ;
default:
return HZ;
}
@ -225,7 +241,7 @@ static unsigned long *pf_peek_vf_state(struct xe_gt *gt, unsigned int vfid)
{
struct xe_gt_sriov_control_state *cs = pf_pick_vf_control(gt, vfid);
return &cs->state;
return cs->state;
}
static bool pf_check_vf_state(struct xe_gt *gt, unsigned int vfid,
@ -329,6 +345,8 @@ static void pf_exit_vf_mismatch(struct xe_gt *gt, unsigned int vfid)
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_FAILED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESUME_FAILED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_FLR_FAILED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_FAILED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_FAILED);
}
#define pf_enter_vf_state_machine_bug(gt, vfid) ({ \
@ -359,6 +377,8 @@ static void pf_queue_vf(struct xe_gt *gt, unsigned int vfid)
static void pf_exit_vf_flr_wip(struct xe_gt *gt, unsigned int vfid);
static void pf_exit_vf_stop_wip(struct xe_gt *gt, unsigned int vfid);
static void pf_exit_vf_save_wip(struct xe_gt *gt, unsigned int vfid);
static void pf_exit_vf_restore_wip(struct xe_gt *gt, unsigned int vfid);
static void pf_exit_vf_pause_wip(struct xe_gt *gt, unsigned int vfid);
static void pf_exit_vf_resume_wip(struct xe_gt *gt, unsigned int vfid);
@ -380,6 +400,8 @@ static void pf_exit_vf_wip(struct xe_gt *gt, unsigned int vfid)
pf_exit_vf_flr_wip(gt, vfid);
pf_exit_vf_stop_wip(gt, vfid);
pf_exit_vf_save_wip(gt, vfid);
pf_exit_vf_restore_wip(gt, vfid);
pf_exit_vf_pause_wip(gt, vfid);
pf_exit_vf_resume_wip(gt, vfid);
@ -399,6 +421,8 @@ static void pf_enter_vf_ready(struct xe_gt *gt, unsigned int vfid)
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_STOPPED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESUMED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORED);
pf_exit_vf_mismatch(gt, vfid);
pf_exit_vf_wip(gt, vfid);
}
@ -429,8 +453,7 @@ static void pf_enter_vf_ready(struct xe_gt *gt, unsigned int vfid)
* : PAUSE_GUC_DONE o-----restart
* : | :
* : | o---<--busy :
* : v / / :
* : PAUSE_SAVE_GUC :
* : / :
* : / :
* : / :
* :....o..............o...............o...........:
@ -450,7 +473,6 @@ static void pf_exit_vf_pause_wip(struct xe_gt *gt, unsigned int vfid)
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_SEND_PAUSE);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_WAIT_GUC);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_GUC_DONE);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_SAVE_GUC);
}
}
@ -481,41 +503,12 @@ static void pf_enter_vf_pause_rejected(struct xe_gt *gt, unsigned int vfid)
pf_enter_vf_pause_failed(gt, vfid);
}
static void pf_enter_vf_pause_save_guc(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_SAVE_GUC))
pf_enter_vf_state_machine_bug(gt, vfid);
}
static bool pf_exit_vf_pause_save_guc(struct xe_gt *gt, unsigned int vfid)
{
int err;
if (!pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_SAVE_GUC))
return false;
err = xe_gt_sriov_pf_migration_save_guc_state(gt, vfid);
if (err) {
/* retry if busy */
if (err == -EBUSY) {
pf_enter_vf_pause_save_guc(gt, vfid);
return true;
}
/* give up on error */
if (err == -EIO)
pf_enter_vf_mismatch(gt, vfid);
}
pf_enter_vf_pause_completed(gt, vfid);
return true;
}
static bool pf_exit_vf_pause_guc_done(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSE_GUC_DONE))
return false;
pf_enter_vf_pause_save_guc(gt, vfid);
pf_enter_vf_pause_completed(gt, vfid);
return true;
}
@ -675,6 +668,8 @@ static void pf_enter_vf_resumed(struct xe_gt *gt, unsigned int vfid)
{
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESUMED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORED);
pf_exit_vf_mismatch(gt, vfid);
pf_exit_vf_wip(gt, vfid);
}
@ -753,6 +748,16 @@ int xe_gt_sriov_pf_control_resume_vf(struct xe_gt *gt, unsigned int vfid)
return -EPERM;
}
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WIP)) {
xe_gt_sriov_dbg(gt, "VF%u save is in progress!\n", vfid);
return -EBUSY;
}
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WIP)) {
xe_gt_sriov_dbg(gt, "VF%u restore is in progress!\n", vfid);
return -EBUSY;
}
if (!pf_enter_vf_resume_wip(gt, vfid)) {
xe_gt_sriov_dbg(gt, "VF%u resume already in progress!\n", vfid);
return -EALREADY;
@ -776,6 +781,562 @@ int xe_gt_sriov_pf_control_resume_vf(struct xe_gt *gt, unsigned int vfid)
return -ECANCELED;
}
/**
* DOC: The VF SAVE state machine
*
* SAVE extends the PAUSED state.
*
* The VF SAVE state machine looks like::
*
* ....PAUSED....................................................
* : :
* : (O)<---------o :
* : | \ :
* : save (SAVED) (SAVE_FAILED) :
* : | ^ ^ :
* : | | | :
* : ....V...............o...........o......SAVE_WIP......... :
* : : | | | : :
* : : | empty | : :
* : : | | | : :
* : : | | | : :
* : : | DATA_DONE | : :
* : : | ^ | : :
* : : | | error : :
* : : | no_data / : :
* : : | / / : :
* : : | / / : :
* : : | / / : :
* : : o---------->PROCESS_DATA<----consume : :
* : : \ \ : :
* : : \ \ : :
* : : \ \ : :
* : : ring_full----->WAIT_DATA : :
* : : : :
* : :......................................................: :
* :............................................................:
*
* For the full state machine view, see `The VF state machine`_.
*/
static void pf_exit_vf_save_wip(struct xe_gt *gt, unsigned int vfid)
{
if (pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WIP)) {
xe_gt_sriov_pf_migration_ring_free(gt, vfid);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WAIT_DATA);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_DATA_DONE);
}
}
static void pf_enter_vf_saved(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVED))
pf_enter_vf_state_machine_bug(gt, vfid);
xe_gt_sriov_dbg(gt, "VF%u saved!\n", vfid);
pf_expect_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED);
pf_exit_vf_mismatch(gt, vfid);
pf_exit_vf_wip(gt, vfid);
}
static void pf_enter_vf_save_failed(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_FAILED))
pf_enter_vf_state_machine_bug(gt, vfid);
wake_up_all(xe_sriov_pf_migration_waitqueue(gt_to_xe(gt), vfid));
pf_exit_vf_wip(gt, vfid);
}
static int pf_handle_vf_save_data(struct xe_gt *gt, unsigned int vfid)
{
int ret;
if (xe_gt_sriov_pf_migration_save_data_pending(gt, vfid,
XE_SRIOV_PACKET_TYPE_GUC)) {
ret = xe_gt_sriov_pf_migration_guc_save(gt, vfid);
if (ret)
return ret;
xe_gt_sriov_pf_migration_save_data_complete(gt, vfid,
XE_SRIOV_PACKET_TYPE_GUC);
return -EAGAIN;
}
if (xe_gt_sriov_pf_migration_save_data_pending(gt, vfid,
XE_SRIOV_PACKET_TYPE_GGTT)) {
ret = xe_gt_sriov_pf_migration_ggtt_save(gt, vfid);
if (ret)
return ret;
xe_gt_sriov_pf_migration_save_data_complete(gt, vfid,
XE_SRIOV_PACKET_TYPE_GGTT);
return -EAGAIN;
}
if (xe_gt_sriov_pf_migration_save_data_pending(gt, vfid,
XE_SRIOV_PACKET_TYPE_MMIO)) {
ret = xe_gt_sriov_pf_migration_mmio_save(gt, vfid);
if (ret)
return ret;
xe_gt_sriov_pf_migration_save_data_complete(gt, vfid,
XE_SRIOV_PACKET_TYPE_MMIO);
return -EAGAIN;
}
if (xe_gt_sriov_pf_migration_save_data_pending(gt, vfid,
XE_SRIOV_PACKET_TYPE_VRAM)) {
ret = xe_gt_sriov_pf_migration_vram_save(gt, vfid);
if (ret == -EAGAIN)
return -EAGAIN;
else if (ret)
return ret;
xe_gt_sriov_pf_migration_save_data_complete(gt, vfid,
XE_SRIOV_PACKET_TYPE_VRAM);
return -EAGAIN;
}
return 0;
}
static bool pf_handle_vf_save(struct xe_gt *gt, unsigned int vfid)
{
int ret;
if (!pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA))
return false;
if (xe_gt_sriov_pf_migration_ring_full(gt, vfid)) {
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WAIT_DATA);
return true;
}
ret = pf_handle_vf_save_data(gt, vfid);
if (ret == -EAGAIN)
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA);
else if (ret)
pf_enter_vf_save_failed(gt, vfid);
else
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_DATA_DONE);
return true;
}
static void pf_exit_vf_save_wait_data(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WAIT_DATA))
return;
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA);
pf_queue_vf(gt, vfid);
}
static bool pf_enter_vf_save_wip(struct xe_gt *gt, unsigned int vfid)
{
if (pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WIP)) {
xe_gt_sriov_pf_migration_save_init(gt, vfid);
pf_enter_vf_wip(gt, vfid);
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA);
pf_queue_vf(gt, vfid);
return true;
}
return false;
}
/**
* xe_gt_sriov_pf_control_check_save_data_done() - Check if all save migration data was produced.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: true if all migration data was produced, false otherwise.
*/
bool xe_gt_sriov_pf_control_check_save_data_done(struct xe_gt *gt, unsigned int vfid)
{
return pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_DATA_DONE);
}
/**
* xe_gt_sriov_pf_control_check_save_failed() - Check if save processing has failed.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: true if save processing failed, false otherwise.
*/
bool xe_gt_sriov_pf_control_check_save_failed(struct xe_gt *gt, unsigned int vfid)
{
return pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_FAILED);
}
/**
* xe_gt_sriov_pf_control_process_save_data() - Queue VF save migration data processing.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_control_process_save_data(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_expect_vf_not_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_FAILED))
return -EIO;
pf_exit_vf_save_wait_data(gt, vfid);
return 0;
}
/**
* xe_gt_sriov_pf_control_trigger_save_vf() - Start an SR-IOV VF migration data save sequence.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_control_trigger_save_vf(struct xe_gt *gt, unsigned int vfid)
{
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_STOPPED)) {
xe_gt_sriov_dbg(gt, "VF%u is stopped!\n", vfid);
return -EPERM;
}
if (!pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED)) {
xe_gt_sriov_dbg(gt, "VF%u is not paused!\n", vfid);
return -EPERM;
}
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WIP)) {
xe_gt_sriov_dbg(gt, "VF%u restore is in progress!\n", vfid);
return -EBUSY;
}
if (!pf_enter_vf_save_wip(gt, vfid)) {
xe_gt_sriov_dbg(gt, "VF%u save already in progress!\n", vfid);
return -EALREADY;
}
return 0;
}
/**
* xe_gt_sriov_pf_control_finish_save_vf() - Complete a VF migration data save sequence.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_control_finish_save_vf(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_DATA_DONE)) {
xe_gt_sriov_err(gt, "VF%u save is still in progress!\n", vfid);
return -EIO;
}
pf_expect_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_DATA_DONE);
pf_enter_vf_saved(gt, vfid);
return 0;
}
/**
* DOC: The VF RESTORE state machine
*
* RESTORE extends the PAUSED state.
*
* The VF RESTORE state machine looks like::
*
* ....PAUSED....................................................
* : :
* : (O)<---------o :
* : | \ :
* : restore (RESTORED) (RESTORE_FAILED) :
* : | ^ ^ :
* : | | | :
* : ....V...............o...........o......RESTORE_WIP...... :
* : : | | | : :
* : : | empty | : :
* : : | | | : :
* : : | | | : :
* : : | DATA_DONE | : :
* : : | ^ | : :
* : : | | error : :
* : : | trailer / : :
* : : | / / : :
* : : | / / : :
* : : | / / : :
* : : o---------->PROCESS_DATA<----produce : :
* : : \ \ : :
* : : \ \ : :
* : : \ \ : :
* : : ring_empty---->WAIT_DATA : :
* : : : :
* : :......................................................: :
* :............................................................:
*
* For the full state machine view, see `The VF state machine`_.
*/
static void pf_exit_vf_restore_wip(struct xe_gt *gt, unsigned int vfid)
{
if (pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WIP)) {
xe_gt_sriov_pf_migration_ring_free(gt, vfid);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_PROCESS_DATA);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WAIT_DATA);
pf_escape_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_DATA_DONE);
}
}
static void pf_enter_vf_restored(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORED))
pf_enter_vf_state_machine_bug(gt, vfid);
xe_gt_sriov_dbg(gt, "VF%u restored!\n", vfid);
pf_expect_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED);
pf_exit_vf_mismatch(gt, vfid);
pf_exit_vf_wip(gt, vfid);
}
static void pf_enter_vf_restore_failed(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_FAILED))
pf_enter_vf_state_machine_bug(gt, vfid);
wake_up_all(xe_sriov_pf_migration_waitqueue(gt_to_xe(gt), vfid));
pf_exit_vf_wip(gt, vfid);
}
static int pf_handle_vf_restore_data(struct xe_gt *gt, unsigned int vfid)
{
struct xe_sriov_packet *data = xe_gt_sriov_pf_migration_restore_consume(gt, vfid);
int ret = 0;
switch (data->hdr.type) {
case XE_SRIOV_PACKET_TYPE_GGTT:
ret = xe_gt_sriov_pf_migration_ggtt_restore(gt, vfid, data);
break;
case XE_SRIOV_PACKET_TYPE_MMIO:
ret = xe_gt_sriov_pf_migration_mmio_restore(gt, vfid, data);
break;
case XE_SRIOV_PACKET_TYPE_GUC:
ret = xe_gt_sriov_pf_migration_guc_restore(gt, vfid, data);
break;
case XE_SRIOV_PACKET_TYPE_VRAM:
ret = xe_gt_sriov_pf_migration_vram_restore(gt, vfid, data);
break;
default:
xe_gt_sriov_notice(gt, "Skipping VF%u unknown data type: %d\n",
vfid, data->hdr.type);
break;
}
xe_sriov_packet_free(data);
return ret;
}
static bool pf_handle_vf_restore(struct xe_gt *gt, unsigned int vfid)
{
int ret;
if (!pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_PROCESS_DATA))
return false;
if (xe_gt_sriov_pf_migration_ring_empty(gt, vfid)) {
if (pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_DATA_DONE))
pf_enter_vf_restored(gt, vfid);
else
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WAIT_DATA);
return true;
}
ret = pf_handle_vf_restore_data(gt, vfid);
if (ret)
pf_enter_vf_restore_failed(gt, vfid);
else
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_PROCESS_DATA);
return true;
}
static void pf_exit_vf_restore_wait_data(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WAIT_DATA))
return;
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_PROCESS_DATA);
pf_queue_vf(gt, vfid);
}
static bool pf_enter_vf_restore_wip(struct xe_gt *gt, unsigned int vfid)
{
if (pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WIP)) {
pf_enter_vf_wip(gt, vfid);
pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_PROCESS_DATA);
pf_queue_vf(gt, vfid);
return true;
}
return false;
}
/**
* xe_gt_sriov_pf_control_check_restore_failed() - Check if restore processing has failed.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: true if restore processing failed, false otherwise.
*/
bool xe_gt_sriov_pf_control_check_restore_failed(struct xe_gt *gt, unsigned int vfid)
{
return pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_FAILED);
}
/**
* xe_gt_sriov_pf_control_restore_data_done() - Indicate the end of VF migration data stream.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_control_restore_data_done(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_DATA_DONE)) {
pf_enter_vf_state_machine_bug(gt, vfid);
return -EIO;
}
return xe_gt_sriov_pf_control_process_restore_data(gt, vfid);
}
/**
* xe_gt_sriov_pf_control_process_restore_data() - Queue VF restore migration data processing.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_control_process_restore_data(struct xe_gt *gt, unsigned int vfid)
{
if (!pf_expect_vf_not_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_FAILED)) {
xe_gt_sriov_pf_migration_ring_free(gt, vfid);
return -EIO;
}
pf_exit_vf_restore_wait_data(gt, vfid);
return 0;
}
/**
* xe_gt_sriov_pf_control_trigger restore_vf() - Start an SR-IOV VF migration data restore sequence.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_control_trigger_restore_vf(struct xe_gt *gt, unsigned int vfid)
{
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_STOPPED)) {
xe_gt_sriov_dbg(gt, "VF%u is stopped!\n", vfid);
return -EPERM;
}
if (!pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED)) {
xe_gt_sriov_dbg(gt, "VF%u is not paused!\n", vfid);
return -EPERM;
}
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WIP)) {
xe_gt_sriov_dbg(gt, "VF%u save is in progress!\n", vfid);
return -EBUSY;
}
if (!pf_enter_vf_restore_wip(gt, vfid)) {
xe_gt_sriov_dbg(gt, "VF%u restore already in progress!\n", vfid);
return -EALREADY;
}
return 0;
}
static int pf_wait_vf_restore_done(struct xe_gt *gt, unsigned int vfid)
{
unsigned long timeout = pf_get_default_timeout(XE_GT_SRIOV_STATE_RESTORE_WIP);
int err;
err = pf_wait_vf_wip_done(gt, vfid, timeout);
if (err) {
xe_gt_sriov_notice(gt, "VF%u RESTORE didn't finish in %u ms (%pe)\n",
vfid, jiffies_to_msecs(timeout), ERR_PTR(err));
return err;
}
if (!pf_expect_vf_not_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_FAILED))
return -EIO;
return 0;
}
/**
* xe_gt_sriov_pf_control_finish_restore_vf() - Complete a VF migration data restore sequence.
* @gt: the &xe_gt
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_gt_sriov_pf_control_finish_restore_vf(struct xe_gt *gt, unsigned int vfid)
{
int ret;
ret = pf_wait_vf_restore_done(gt, vfid);
if (ret)
return ret;
if (!pf_expect_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORED)) {
pf_enter_vf_mismatch(gt, vfid);
return -EIO;
}
pf_expect_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED);
return 0;
}
/**
* DOC: The VF STOP state machine
*
@ -817,6 +1378,8 @@ static void pf_enter_vf_stopped(struct xe_gt *gt, unsigned int vfid)
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESUMED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_PAUSED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVED);
pf_exit_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORED);
pf_exit_vf_mismatch(gt, vfid);
pf_exit_vf_wip(gt, vfid);
}
@ -1460,7 +2023,22 @@ static bool pf_process_vf_state_machine(struct xe_gt *gt, unsigned int vfid)
if (pf_exit_vf_pause_guc_done(gt, vfid))
return true;
if (pf_exit_vf_pause_save_guc(gt, vfid))
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WAIT_DATA)) {
xe_gt_sriov_dbg_verbose(gt, "VF%u in %s\n", vfid,
control_bit_to_string(XE_GT_SRIOV_STATE_SAVE_WAIT_DATA));
return false;
}
if (pf_handle_vf_save(gt, vfid))
return true;
if (pf_check_vf_state(gt, vfid, XE_GT_SRIOV_STATE_RESTORE_WAIT_DATA)) {
xe_gt_sriov_dbg_verbose(gt, "VF%u in %s\n", vfid,
control_bit_to_string(XE_GT_SRIOV_STATE_RESTORE_WAIT_DATA));
return false;
}
if (pf_handle_vf_restore(gt, vfid))
return true;
if (pf_exit_vf_resume_send_resume(gt, vfid))

10
drivers/gpu/drm/xe/xe_gt_sriov_pf_control.h
View File

@ -16,6 +16,16 @@ void xe_gt_sriov_pf_control_restart(struct xe_gt *gt);
int xe_gt_sriov_pf_control_pause_vf(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_resume_vf(struct xe_gt *gt, unsigned int vfid);
bool xe_gt_sriov_pf_control_check_save_data_done(struct xe_gt *gt, unsigned int vfid);
bool xe_gt_sriov_pf_control_check_save_failed(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_process_save_data(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_trigger_save_vf(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_finish_save_vf(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_restore_data_done(struct xe_gt *gt, unsigned int vfid);
bool xe_gt_sriov_pf_control_check_restore_failed(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_process_restore_data(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_trigger_restore_vf(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_finish_restore_vf(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_stop_vf(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_trigger_flr(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_control_sync_flr(struct xe_gt *gt, unsigned int vfid, bool sync);

34
drivers/gpu/drm/xe/xe_gt_sriov_pf_control_types.h
View File

@ -28,9 +28,20 @@
* @XE_GT_SRIOV_STATE_PAUSE_SEND_PAUSE: indicates that the PF is about to send a PAUSE command.
* @XE_GT_SRIOV_STATE_PAUSE_WAIT_GUC: indicates that the PF awaits for a response from the GuC.
* @XE_GT_SRIOV_STATE_PAUSE_GUC_DONE: indicates that the PF has received a response from the GuC.
* @XE_GT_SRIOV_STATE_PAUSE_SAVE_GUC: indicates that the PF needs to save the VF GuC state.
* @XE_GT_SRIOV_STATE_PAUSE_FAILED: indicates that a VF pause operation has failed.
* @XE_GT_SRIOV_STATE_PAUSED: indicates that the VF is paused.
* @XE_GT_SRIOV_STATE_SAVE_WIP: indicates that VF save operation is in progress.
* @XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA: indicates that VF migration data is being produced.
* @XE_GT_SRIOV_STATE_SAVE_WAIT_DATA: indicates that PF awaits for space in migration data ring.
* @XE_GT_SRIOV_STATE_SAVE_DATA_DONE: indicates that all migration data was produced by Xe.
* @XE_GT_SRIOV_STATE_SAVE_FAILED: indicates that VF save operation has failed.
* @XE_GT_SRIOV_STATE_SAVED: indicates that VF data is saved.
* @XE_GT_SRIOV_STATE_RESTORE_WIP: indicates that VF restore operation is in progress.
* @XE_GT_SRIOV_STATE_RESTORE_PROCESS_DATA: indicates that VF migration data is being consumed.
* @XE_GT_SRIOV_STATE_RESTORE_WAIT_DATA: indicates that PF awaits for data in migration data ring.
* @XE_GT_SRIOV_STATE_RESTORE_DATA_DONE: indicates that all migration data was produced by the user.
* @XE_GT_SRIOV_STATE_RESTORE_FAILED: indicates that VF restore operation has failed.
* @XE_GT_SRIOV_STATE_RESTORED: indicates that VF data is restored.
* @XE_GT_SRIOV_STATE_RESUME_WIP: indicates the a VF resume operation is in progress.
* @XE_GT_SRIOV_STATE_RESUME_SEND_RESUME: indicates that the PF is about to send RESUME command.
* @XE_GT_SRIOV_STATE_RESUME_FAILED: indicates that a VF resume operation has failed.
@ -59,10 +70,23 @@ enum xe_gt_sriov_control_bits {
XE_GT_SRIOV_STATE_PAUSE_SEND_PAUSE,
XE_GT_SRIOV_STATE_PAUSE_WAIT_GUC,
XE_GT_SRIOV_STATE_PAUSE_GUC_DONE,
XE_GT_SRIOV_STATE_PAUSE_SAVE_GUC,
XE_GT_SRIOV_STATE_PAUSE_FAILED,
XE_GT_SRIOV_STATE_PAUSED,
XE_GT_SRIOV_STATE_SAVE_WIP,
XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA,
XE_GT_SRIOV_STATE_SAVE_WAIT_DATA,
XE_GT_SRIOV_STATE_SAVE_DATA_DONE,
XE_GT_SRIOV_STATE_SAVE_FAILED,
XE_GT_SRIOV_STATE_SAVED,
XE_GT_SRIOV_STATE_RESTORE_WIP,
XE_GT_SRIOV_STATE_RESTORE_PROCESS_DATA,
XE_GT_SRIOV_STATE_RESTORE_WAIT_DATA,
XE_GT_SRIOV_STATE_RESTORE_DATA_DONE,
XE_GT_SRIOV_STATE_RESTORE_FAILED,
XE_GT_SRIOV_STATE_RESTORED,
XE_GT_SRIOV_STATE_RESUME_WIP,
XE_GT_SRIOV_STATE_RESUME_SEND_RESUME,
XE_GT_SRIOV_STATE_RESUME_FAILED,
@ -73,9 +97,11 @@ enum xe_gt_sriov_control_bits {
XE_GT_SRIOV_STATE_STOP_FAILED,
XE_GT_SRIOV_STATE_STOPPED,
XE_GT_SRIOV_STATE_MISMATCH = BITS_PER_LONG - 1,
XE_GT_SRIOV_STATE_MISMATCH, /* always keep as last */
};
#define XE_GT_SRIOV_NUM_STATES (XE_GT_SRIOV_STATE_MISMATCH + 1)
/**
* struct xe_gt_sriov_control_state - GT-level per-VF control state.
*
@ -83,7 +109,7 @@ enum xe_gt_sriov_control_bits {
*/
struct xe_gt_sriov_control_state {
/** @state: VF state bits */
unsigned long state;
DECLARE_BITMAP(state, XE_GT_SRIOV_NUM_STATES);
/** @done: completion of async operations */
struct completion done;

47
drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.c
View File

@ -327,9 +327,6 @@ static const struct {
{ "stop", xe_gt_sriov_pf_control_stop_vf },
{ "pause", xe_gt_sriov_pf_control_pause_vf },
{ "resume", xe_gt_sriov_pf_control_resume_vf },
#ifdef CONFIG_DRM_XE_DEBUG_SRIOV
{ "restore!", xe_gt_sriov_pf_migration_restore_guc_state },
#endif
};
static ssize_t control_write(struct file *file, const char __user *buf, size_t count, loff_t *pos)
@ -393,47 +390,6 @@ static const struct file_operations control_ops = {
.llseek = default_llseek,
};
/*
* /sys/kernel/debug/dri/BDF/
* ├── sriov
* : ├── vf1
* : ├── tile0
* : ├── gt0
* : ├── guc_state
*/
static ssize_t guc_state_read(struct file *file, char __user *buf,
size_t count, loff_t *pos)
{
struct dentry *dent = file_dentry(file);
struct dentry *parent = dent->d_parent;
struct xe_gt *gt = extract_gt(parent);
unsigned int vfid = extract_vfid(parent);
return xe_gt_sriov_pf_migration_read_guc_state(gt, vfid, buf, count, pos);
}
static ssize_t guc_state_write(struct file *file, const char __user *buf,
size_t count, loff_t *pos)
{
struct dentry *dent = file_dentry(file);
struct dentry *parent = dent->d_parent;
struct xe_gt *gt = extract_gt(parent);
unsigned int vfid = extract_vfid(parent);
if (*pos)
return -EINVAL;
return xe_gt_sriov_pf_migration_write_guc_state(gt, vfid, buf, count);
}
static const struct file_operations guc_state_ops = {
.owner = THIS_MODULE,
.read = guc_state_read,
.write = guc_state_write,
.llseek = default_llseek,
};
/*
* /sys/kernel/debug/dri/BDF/
* ├── sriov
@ -568,9 +524,6 @@ static void pf_populate_gt(struct xe_gt *gt, struct dentry *dent, unsigned int v
/* for testing/debugging purposes only! */
if (IS_ENABLED(CONFIG_DRM_XE_DEBUG)) {
debugfs_create_file("guc_state",
IS_ENABLED(CONFIG_DRM_XE_DEBUG_SRIOV) ? 0600 : 0400,
dent, NULL, &guc_state_ops);
debugfs_create_file("config_blob",
IS_ENABLED(CONFIG_DRM_XE_DEBUG_SRIOV) ? 0600 : 0400,
dent, NULL, &config_blob_ops);

1112
drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c
View File

File diff suppressed because it is too large Load Diff

46
drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.h
View File

@ -9,16 +9,46 @@
#include <linux/types.h>
struct xe_gt;
struct xe_sriov_packet;
enum xe_sriov_packet_type;
/* TODO: get this information by querying GuC in the future */
#define XE_GT_SRIOV_PF_MIGRATION_GUC_DATA_MAX_SIZE SZ_8M
int xe_gt_sriov_pf_migration_init(struct xe_gt *gt);
int xe_gt_sriov_pf_migration_save_guc_state(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_migration_restore_guc_state(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_migration_guc_save(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_migration_guc_restore(struct xe_gt *gt, unsigned int vfid,
struct xe_sriov_packet *data);
int xe_gt_sriov_pf_migration_ggtt_save(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_migration_ggtt_restore(struct xe_gt *gt, unsigned int vfid,
struct xe_sriov_packet *data);
int xe_gt_sriov_pf_migration_mmio_save(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_migration_mmio_restore(struct xe_gt *gt, unsigned int vfid,
struct xe_sriov_packet *data);
int xe_gt_sriov_pf_migration_vram_save(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_migration_vram_restore(struct xe_gt *gt, unsigned int vfid,
struct xe_sriov_packet *data);
#ifdef CONFIG_DEBUG_FS
ssize_t xe_gt_sriov_pf_migration_read_guc_state(struct xe_gt *gt, unsigned int vfid,
char __user *buf, size_t count, loff_t *pos);
ssize_t xe_gt_sriov_pf_migration_write_guc_state(struct xe_gt *gt, unsigned int vfid,
const char __user *buf, size_t count);
#endif
ssize_t xe_gt_sriov_pf_migration_size(struct xe_gt *gt, unsigned int vfid);
bool xe_gt_sriov_pf_migration_ring_empty(struct xe_gt *gt, unsigned int vfid);
bool xe_gt_sriov_pf_migration_ring_full(struct xe_gt *gt, unsigned int vfid);
void xe_gt_sriov_pf_migration_ring_free(struct xe_gt *gt, unsigned int vfid);
void xe_gt_sriov_pf_migration_save_init(struct xe_gt *gt, unsigned int vfid);
bool xe_gt_sriov_pf_migration_save_data_pending(struct xe_gt *gt, unsigned int vfid,
enum xe_sriov_packet_type type);
void xe_gt_sriov_pf_migration_save_data_complete(struct xe_gt *gt, unsigned int vfid,
enum xe_sriov_packet_type type);
int xe_gt_sriov_pf_migration_save_produce(struct xe_gt *gt, unsigned int vfid,
struct xe_sriov_packet *data);
struct xe_sriov_packet *
xe_gt_sriov_pf_migration_restore_consume(struct xe_gt *gt, unsigned int vfid);
int xe_gt_sriov_pf_migration_restore_produce(struct xe_gt *gt, unsigned int vfid,
struct xe_sriov_packet *data);
struct xe_sriov_packet *
xe_gt_sriov_pf_migration_save_consume(struct xe_gt *gt, unsigned int vfid);
#endif

34
drivers/gpu/drm/xe/xe_gt_sriov_pf_migration_types.h
View File

@ -6,35 +6,23 @@
#ifndef _XE_GT_SRIOV_PF_MIGRATION_TYPES_H_
#define _XE_GT_SRIOV_PF_MIGRATION_TYPES_H_
#include <linux/mutex.h>
#include <linux/types.h>
#include <linux/ptr_ring.h>
/**
* struct xe_gt_sriov_state_snapshot - GT-level per-VF state snapshot data.
* struct xe_gt_sriov_migration_data - GT-level per-VF migration data.
*
* Used by the PF driver to maintain per-VF migration data.
*/
struct xe_gt_sriov_state_snapshot {
/** @guc: GuC VF state snapshot */
struct xe_gt_sriov_migration_data {
/** @ring: queue containing VF save / restore migration data */
struct ptr_ring ring;
/** @save: structure for currently processed save migration data */
struct {
/** @guc.buff: buffer with the VF state */
u32 *buff;
/** @guc.size: size of the buffer (must be dwords aligned) */
u32 size;
} guc;
};
/**
* struct xe_gt_sriov_pf_migration - GT-level data.
*
* Used by the PF driver to maintain non-VF specific per-GT data.
*/
struct xe_gt_sriov_pf_migration {
/** @supported: indicates whether the feature is supported */
bool supported;
/** @snapshot_lock: protects all VFs snapshots */
struct mutex snapshot_lock;
/** @save.data_remaining: bitmap of migration types that need to be saved */
unsigned long data_remaining;
/** @save.vram_offset: last saved offset within VRAM, used for chunked VRAM save */
loff_t vram_offset;
} save;
};
#endif

21
drivers/gpu/drm/xe/xe_gt_sriov_pf_service.c
View File

@ -99,11 +99,30 @@ static const struct xe_reg ver_3000_runtime_regs[] = {
HUC_KERNEL_LOAD_INFO, /* _MMIO(0xc1dc) */
};
static const struct xe_reg ver_35_runtime_regs[] = {
RPM_CONFIG0, /* _MMIO(0x0d00) */
XEHP_FUSE4, /* _MMIO(0x9114) */
MIRROR_FUSE3, /* _MMIO(0x9118) */
MIRROR_L3BANK_ENABLE, /* _MMIO(0x9130) */
XELP_EU_ENABLE, /* _MMIO(0x9134) */
XELP_GT_GEOMETRY_DSS_ENABLE, /* _MMIO(0x913c) */
GT_VEBOX_VDBOX_DISABLE, /* _MMIO(0x9140) */
XEHP_GT_COMPUTE_DSS_ENABLE, /* _MMIO(0x9144) */
XEHPC_GT_COMPUTE_DSS_ENABLE_EXT,/* _MMIO(0x9148) */
XE2_GT_COMPUTE_DSS_2, /* _MMIO(0x914c) */
XE2_GT_GEOMETRY_DSS_1, /* _MMIO(0x9150) */
XE2_GT_GEOMETRY_DSS_2, /* _MMIO(0x9154) */
SERVICE_COPY_ENABLE, /* _MMIO(0x9170) */
};
static const struct xe_reg *pick_runtime_regs(struct xe_device *xe, unsigned int *count)
{
const struct xe_reg *regs;
if (GRAPHICS_VERx100(xe) >= 3000) {
if (GRAPHICS_VER(xe) >= 35) {
*count = ARRAY_SIZE(ver_35_runtime_regs);
regs = ver_35_runtime_regs;
} else if (GRAPHICS_VERx100(xe) >= 3000) {
*count = ARRAY_SIZE(ver_3000_runtime_regs);
regs = ver_3000_runtime_regs;
} else if (GRAPHICS_VERx100(xe) >= 2000) {

5
drivers/gpu/drm/xe/xe_gt_sriov_pf_types.h
View File

@ -31,8 +31,8 @@ struct xe_gt_sriov_metadata {
/** @version: negotiated VF/PF ABI version */
struct xe_gt_sriov_pf_service_version version;
/** @snapshot: snapshot of the VF state data */
struct xe_gt_sriov_state_snapshot snapshot;
/** @migration: per-VF migration data. */
struct xe_gt_sriov_migration_data migration;
};
/**
@ -58,7 +58,6 @@ struct xe_gt_sriov_pf {
struct xe_gt_sriov_pf_service service;
struct xe_gt_sriov_pf_control control;
struct xe_gt_sriov_pf_policy policy;
struct xe_gt_sriov_pf_migration migration;
struct xe_gt_sriov_spare_config spare;
struct xe_gt_sriov_metadata *vfs;
};

58
drivers/gpu/drm/xe/xe_gt_throttle.c
View File

@ -22,9 +22,15 @@
* Their availability depend on the platform and some may not be visible if that
* reason is not available.
*
* The ``reasons`` attribute can be used by sysadmin to monitor all possible
* reasons for throttling and report them. It's preferred over monitoring
* ``status`` and then reading the reason from individual attributes since that
* is racy. If there's no throttling happening, "none" is returned.
*
* The following attributes are available on Crescent Island platform:
*
* - ``status``: Overall throttle status
* - ``status``: Overall throttle status (0: no throttling, 1: throttling)
* - ``reasons``: Array of reasons causing throttling separated by space
* - ``reason_pl1``: package PL1
* - ``reason_pl2``: package PL2
* - ``reason_pl4``: package PL4
@ -43,7 +49,8 @@
*
* Other platforms support the following reasons:
*
* - ``status``: Overall status
* - ``status``: Overall throttle status (0: no throttling, 1: throttling)
* - ``reasons``: Array of reasons causing throttling separated by space
* - ``reason_pl1``: package PL1
* - ``reason_pl2``: package PL2
* - ``reason_pl4``: package PL4, Iccmax etc.
@ -111,12 +118,57 @@ static ssize_t reason_show(struct kobject *kobj,
return sysfs_emit(buff, "%u\n", is_throttled_by(gt, ta->mask));
}
static const struct attribute_group *get_platform_throttle_group(struct xe_device *xe);
static ssize_t reasons_show(struct kobject *kobj,
struct kobj_attribute *attr, char *buff)
{
struct xe_gt *gt = throttle_to_gt(kobj);
struct xe_device *xe = gt_to_xe(gt);
const struct attribute_group *group;
struct attribute **pother;
ssize_t ret = 0;
u32 reasons;
reasons = xe_gt_throttle_get_limit_reasons(gt);
if (!reasons)
goto ret_none;
group = get_platform_throttle_group(xe);
for (pother = group->attrs; *pother; pother++) {
struct kobj_attribute *kattr = container_of(*pother, struct kobj_attribute, attr);
struct throttle_attribute *other_ta = kobj_attribute_to_throttle(kattr);
if (other_ta->mask != U32_MAX && reasons & other_ta->mask)
ret += sysfs_emit_at(buff, ret, "%s ", (*pother)->name);
}
if (drm_WARN_ONCE(&xe->drm, !ret, "Unknown reason: %#x\n", reasons))
goto ret_none;
/* Drop extra space from last iteration above */
ret--;
ret += sysfs_emit_at(buff, ret, "\n");
return ret;
ret_none:
return sysfs_emit(buff, "none\n");
}
#define THROTTLE_ATTR_RO(name, _mask) \
struct throttle_attribute attr_##name = { \
.attr = __ATTR(name, 0444, reason_show, NULL), \
.mask = _mask, \
}
#define THROTTLE_ATTR_RO_FUNC(name, _mask, _show) \
struct throttle_attribute attr_##name = { \
.attr = __ATTR(name, 0444, _show, NULL), \
.mask = _mask, \
}
static THROTTLE_ATTR_RO_FUNC(reasons, 0, reasons_show);
static THROTTLE_ATTR_RO(status, U32_MAX);
static THROTTLE_ATTR_RO(reason_pl1, POWER_LIMIT_1_MASK);
static THROTTLE_ATTR_RO(reason_pl2, POWER_LIMIT_2_MASK);
@ -128,6 +180,7 @@ static THROTTLE_ATTR_RO(reason_vr_thermalert, VR_THERMALERT_MASK);
static THROTTLE_ATTR_RO(reason_vr_tdc, VR_TDC_MASK);
static struct attribute *throttle_attrs[] = {
&attr_reasons.attr.attr,
&attr_status.attr.attr,
&attr_reason_pl1.attr.attr,
&attr_reason_pl2.attr.attr,
@ -153,6 +206,7 @@ static THROTTLE_ATTR_RO(reason_psys_crit, PSYS_CRIT_MASK);
static struct attribute *cri_throttle_attrs[] = {
/* Common */
&attr_reasons.attr.attr,
&attr_status.attr.attr,
&attr_reason_pl1.attr.attr,
&attr_reason_pl2.attr.attr,

119
drivers/gpu/drm/xe/xe_guard.h Normal file
View File

@ -0,0 +1,119 @@
/* SPDX-License-Identifier: MIT */
/*
* Copyright © 2025 Intel Corporation
*/
#ifndef _XE_GUARD_H_
#define _XE_GUARD_H_
#include <linux/spinlock.h>
/**
* struct xe_guard - Simple logic to protect a feature.
*
* Implements simple semaphore-like logic that can be used to lockdown the
* feature unless it is already in use. Allows enabling of the otherwise
* incompatible features, where we can't follow the strict owner semantics
* required by the &rw_semaphore.
*
* NOTE! It shouldn't be used to protect a data, use &rw_semaphore instead.
*/
struct xe_guard {
/**
* @counter: implements simple exclusive/lockdown logic:
* if == 0 then guard/feature is idle/not in use,
* if < 0 then feature is active and can't be locked-down,
* if > 0 then feature is lockded-down and can't be activated.
*/
int counter;
/** @name: the name of the guard (useful for debug) */
const char *name;
/** @owner: the info about the last owner of the guard (for debug) */
void *owner;
/** @lock: protects guard's data */
spinlock_t lock;
};
/**
* xe_guard_init() - Initialize the guard.
* @guard: the &xe_guard to init
* @name: name of the guard
*/
static inline void xe_guard_init(struct xe_guard *guard, const char *name)
{
spin_lock_init(&guard->lock);
guard->counter = 0;
guard->name = name;
}
/**
* xe_guard_arm() - Arm the guard for the exclusive/lockdown mode.
* @guard: the &xe_guard to arm
* @lockdown: arm for lockdown(true) or exclusive(false) mode
* @who: optional owner info (for debug only)
*
* Multiple lockdown requests are allowed.
* Only single exclusive access can be granted.
* Will fail if the guard is already in exclusive mode.
* On success, must call the xe_guard_disarm() to release.
*
* Return: 0 on success or a negative error code on failure.
*/
static inline int xe_guard_arm(struct xe_guard *guard, bool lockdown, void *who)
{
guard(spinlock)(&guard->lock);
if (lockdown) {
if (guard->counter < 0)
return -EBUSY;
guard->counter++;
} else {
if (guard->counter > 0)
return -EPERM;
if (guard->counter < 0)
return -EUSERS;
guard->counter--;
}
guard->owner = who;
return 0;
}
/**
* xe_guard_disarm() - Disarm the guard from exclusive/lockdown mode.
* @guard: the &xe_guard to disarm
* @lockdown: disarm from lockdown(true) or exclusive(false) mode
*
* Return: true if successfully disarmed or false in case of mismatch.
*/
static inline bool xe_guard_disarm(struct xe_guard *guard, bool lockdown)
{
guard(spinlock)(&guard->lock);
if (lockdown) {
if (guard->counter <= 0)
return false;
guard->counter--;
} else {
if (guard->counter != -1)
return false;
guard->counter++;
}
return true;
}
/**
* xe_guard_mode_str() - Convert guard mode into a string.
* @lockdown: flag used to select lockdown or exclusive mode
*
* Return: "lockdown" or "exclusive" string.
*/
static inline const char *xe_guard_mode_str(bool lockdown)
{
return lockdown ? "lockdown" : "exclusive";
}
#endif

13
drivers/gpu/drm/xe/xe_guc.c
View File

@ -24,6 +24,7 @@
#include "xe_gt_printk.h"
#include "xe_gt_sriov_vf.h"
#include "xe_gt_throttle.h"
#include "xe_gt_sriov_pf_migration.h"
#include "xe_guc_ads.h"
#include "xe_guc_buf.h"
#include "xe_guc_capture.h"
@ -40,6 +41,7 @@
#include "xe_mmio.h"
#include "xe_platform_types.h"
#include "xe_sriov.h"
#include "xe_sriov_pf_migration.h"
#include "xe_uc.h"
#include "xe_uc_fw.h"
#include "xe_wa.h"
@ -821,6 +823,14 @@ static int vf_guc_init_post_hwconfig(struct xe_guc *guc)
return 0;
}
static u32 guc_additional_cache_size(struct xe_device *xe)
{
if (IS_SRIOV_PF(xe) && xe_sriov_pf_migration_supported(xe))
return XE_GT_SRIOV_PF_MIGRATION_GUC_DATA_MAX_SIZE;
else
return 0; /* Fallback to default size */
}
/**
* xe_guc_init_post_hwconfig - initialize GuC post hwconfig load
* @guc: The GuC object
@ -860,7 +870,8 @@ int xe_guc_init_post_hwconfig(struct xe_guc *guc)
if (ret)
return ret;
ret = xe_guc_buf_cache_init(&guc->buf);
ret = xe_guc_buf_cache_init_with_size(&guc->buf,
guc_additional_cache_size(guc_to_xe(guc)));
if (ret)
return ret;

57
drivers/gpu/drm/xe/xe_guc_buf.c
View File

@ -13,6 +13,8 @@
#include "xe_guc_buf.h"
#include "xe_sa.h"
#define XE_GUC_BUF_CACHE_DEFAULT_SIZE SZ_8K
static struct xe_guc *cache_to_guc(struct xe_guc_buf_cache *cache)
{
return container_of(cache, struct xe_guc, buf);
@ -23,21 +25,12 @@ static struct xe_gt *cache_to_gt(struct xe_guc_buf_cache *cache)
return guc_to_gt(cache_to_guc(cache));
}
/**
* xe_guc_buf_cache_init() - Initialize the GuC Buffer Cache.
* @cache: the &xe_guc_buf_cache to initialize
*
* The Buffer Cache allows to obtain a reusable buffer that can be used to pass
* indirect H2G data to GuC without a need to create a ad-hoc allocation.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_guc_buf_cache_init(struct xe_guc_buf_cache *cache)
static int guc_buf_cache_init(struct xe_guc_buf_cache *cache, u32 size)
{
struct xe_gt *gt = cache_to_gt(cache);
struct xe_sa_manager *sam;
sam = __xe_sa_bo_manager_init(gt_to_tile(gt), SZ_8K, 0, sizeof(u32));
sam = __xe_sa_bo_manager_init(gt_to_tile(gt), size, 0, sizeof(u32));
if (IS_ERR(sam))
return PTR_ERR(sam);
cache->sam = sam;
@ -48,6 +41,35 @@ int xe_guc_buf_cache_init(struct xe_guc_buf_cache *cache)
return 0;
}
/**
* xe_guc_buf_cache_init() - Initialize the GuC Buffer Cache.
* @cache: the &xe_guc_buf_cache to initialize
*
* The Buffer Cache allows to obtain a reusable buffer that can be used to pass
* data to GuC or read data from GuC without a need to create a ad-hoc allocation.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_guc_buf_cache_init(struct xe_guc_buf_cache *cache)
{
return guc_buf_cache_init(cache, XE_GUC_BUF_CACHE_DEFAULT_SIZE);
}
/**
* xe_guc_buf_cache_init_with_size() - Initialize the GuC Buffer Cache.
* @cache: the &xe_guc_buf_cache to initialize
* @size: size in bytes
*
* Like xe_guc_buf_cache_init(), except it allows the caller to make the cache
* buffer larger, allowing to accommodate larger objects.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_guc_buf_cache_init_with_size(struct xe_guc_buf_cache *cache, u32 size)
{
return guc_buf_cache_init(cache, max(XE_GUC_BUF_CACHE_DEFAULT_SIZE, size));
}
/**
* xe_guc_buf_cache_dwords() - Number of dwords the GuC Buffer Cache supports.
* @cache: the &xe_guc_buf_cache to query
@ -115,6 +137,19 @@ void xe_guc_buf_release(const struct xe_guc_buf buf)
xe_sa_bo_free(buf.sa, NULL);
}
/**
* xe_guc_buf_sync_read() - Copy the data from the GPU memory to the sub-allocation.
* @buf: the &xe_guc_buf to sync
*
* Return: a CPU pointer of the sub-allocation.
*/
void *xe_guc_buf_sync_read(const struct xe_guc_buf buf)
{
xe_sa_bo_sync_read(buf.sa);
return xe_sa_bo_cpu_addr(buf.sa);
}
/**
* xe_guc_buf_flush() - Copy the data from the sub-allocation to the GPU memory.
* @buf: the &xe_guc_buf to flush

2
drivers/gpu/drm/xe/xe_guc_buf.h
View File

@ -12,6 +12,7 @@
#include "xe_guc_buf_types.h"
int xe_guc_buf_cache_init(struct xe_guc_buf_cache *cache);
int xe_guc_buf_cache_init_with_size(struct xe_guc_buf_cache *cache, u32 size);
u32 xe_guc_buf_cache_dwords(struct xe_guc_buf_cache *cache);
struct xe_guc_buf xe_guc_buf_reserve(struct xe_guc_buf_cache *cache, u32 dwords);
struct xe_guc_buf xe_guc_buf_from_data(struct xe_guc_buf_cache *cache,
@ -30,6 +31,7 @@ static inline bool xe_guc_buf_is_valid(const struct xe_guc_buf buf)
}
void *xe_guc_buf_cpu_ptr(const struct xe_guc_buf buf);
void *xe_guc_buf_sync_read(const struct xe_guc_buf buf);
u64 xe_guc_buf_flush(const struct xe_guc_buf buf);
u64 xe_guc_buf_gpu_addr(const struct xe_guc_buf buf);
u64 xe_guc_cache_gpu_addr_from_ptr(struct xe_guc_buf_cache *cache, const void *ptr, u32 size);

100
drivers/gpu/drm/xe/xe_guc_pc.c
View File

@ -331,7 +331,7 @@ static int pc_set_min_freq(struct xe_guc_pc *pc, u32 freq)
* Our goal is to have the admin choices respected.
*/
pc_action_set_param(pc, SLPC_PARAM_IGNORE_EFFICIENT_FREQUENCY,
freq < pc->rpe_freq);
freq < xe_guc_pc_get_rpe_freq(pc));
return pc_action_set_param(pc,
SLPC_PARAM_GLOBAL_MIN_GT_UNSLICE_FREQ_MHZ,
@ -363,7 +363,7 @@ static int pc_set_max_freq(struct xe_guc_pc *pc, u32 freq)
freq);
}
static void mtl_update_rpa_value(struct xe_guc_pc *pc)
static u32 mtl_get_rpa_freq(struct xe_guc_pc *pc)
{
struct xe_gt *gt = pc_to_gt(pc);
u32 reg;
@ -373,10 +373,10 @@ static void mtl_update_rpa_value(struct xe_guc_pc *pc)
else
reg = xe_mmio_read32(&gt->mmio, MTL_GT_RPA_FREQUENCY);
pc->rpa_freq = decode_freq(REG_FIELD_GET(MTL_RPA_MASK, reg));
return decode_freq(REG_FIELD_GET(MTL_RPA_MASK, reg));
}
static void mtl_update_rpe_value(struct xe_guc_pc *pc)
static u32 mtl_get_rpe_freq(struct xe_guc_pc *pc)
{
struct xe_gt *gt = pc_to_gt(pc);
u32 reg;
@ -386,68 +386,56 @@ static void mtl_update_rpe_value(struct xe_guc_pc *pc)
else
reg = xe_mmio_read32(&gt->mmio, MTL_GT_RPE_FREQUENCY);
pc->rpe_freq = decode_freq(REG_FIELD_GET(MTL_RPE_MASK, reg));
return decode_freq(REG_FIELD_GET(MTL_RPE_MASK, reg));
}
static void tgl_update_rpa_value(struct xe_guc_pc *pc)
static u32 pvc_get_rpa_freq(struct xe_guc_pc *pc)
{
struct xe_gt *gt = pc_to_gt(pc);
struct xe_device *xe = gt_to_xe(gt);
u32 reg;
/*
* For PVC we still need to use fused RP0 as the approximation for RPa
* For other platforms than PVC we get the resolved RPa directly from
* PCODE at a different register
*/
if (xe->info.platform == XE_PVC) {
reg = xe_mmio_read32(&gt->mmio, PVC_RP_STATE_CAP);
pc->rpa_freq = REG_FIELD_GET(RP0_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
} else {
reg = xe_mmio_read32(&gt->mmio, FREQ_INFO_REC);
pc->rpa_freq = REG_FIELD_GET(RPA_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
}
struct xe_gt *gt = pc_to_gt(pc);
u32 reg;
reg = xe_mmio_read32(&gt->mmio, PVC_RP_STATE_CAP);
return REG_FIELD_GET(RP0_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
}
static void tgl_update_rpe_value(struct xe_guc_pc *pc)
static u32 tgl_get_rpa_freq(struct xe_guc_pc *pc)
{
struct xe_gt *gt = pc_to_gt(pc);
u32 reg;
reg = xe_mmio_read32(&gt->mmio, FREQ_INFO_REC);
return REG_FIELD_GET(RPA_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
}
static u32 pvc_get_rpe_freq(struct xe_guc_pc *pc)
{
struct xe_gt *gt = pc_to_gt(pc);
struct xe_device *xe = gt_to_xe(gt);
u32 reg;
/*
* For PVC we still need to use fused RP1 as the approximation for RPe
* For other platforms than PVC we get the resolved RPe directly from
* PCODE at a different register
*/
if (xe->info.platform == XE_PVC) {
reg = xe_mmio_read32(&gt->mmio, PVC_RP_STATE_CAP);
pc->rpe_freq = REG_FIELD_GET(RP1_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
} else {
reg = xe_mmio_read32(&gt->mmio, FREQ_INFO_REC);
pc->rpe_freq = REG_FIELD_GET(RPE_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
}
reg = xe_mmio_read32(&gt->mmio, PVC_RP_STATE_CAP);
return REG_FIELD_GET(RP1_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
}
static void pc_update_rp_values(struct xe_guc_pc *pc)
static u32 tgl_get_rpe_freq(struct xe_guc_pc *pc)
{
struct xe_gt *gt = pc_to_gt(pc);
struct xe_device *xe = gt_to_xe(gt);
if (GRAPHICS_VERx100(xe) >= 1270) {
mtl_update_rpa_value(pc);
mtl_update_rpe_value(pc);
} else {
tgl_update_rpa_value(pc);
tgl_update_rpe_value(pc);
}
u32 reg;
/*
* RPe is decided at runtime by PCODE. In the rare case where that's
* smaller than the fused min, we will trust the PCODE and use that
* as our minimum one.
* For other platforms than PVC, we get the resolved RPe directly from
* PCODE at a different register
*/
pc->rpn_freq = min(pc->rpn_freq, pc->rpe_freq);
reg = xe_mmio_read32(&gt->mmio, FREQ_INFO_REC);
return REG_FIELD_GET(RPE_MASK, reg) * GT_FREQUENCY_MULTIPLIER;
}
/**
@ -548,9 +536,15 @@ u32 xe_guc_pc_get_rp0_freq(struct xe_guc_pc *pc)
*/
u32 xe_guc_pc_get_rpa_freq(struct xe_guc_pc *pc)
{
pc_update_rp_values(pc);
struct xe_gt *gt = pc_to_gt(pc);
struct xe_device *xe = gt_to_xe(gt);
return pc->rpa_freq;
if (GRAPHICS_VERx100(xe) == 1260)
return pvc_get_rpa_freq(pc);
else if (GRAPHICS_VERx100(xe) >= 1270)
return mtl_get_rpa_freq(pc);
else
return tgl_get_rpa_freq(pc);
}
/**
@ -561,9 +555,17 @@ u32 xe_guc_pc_get_rpa_freq(struct xe_guc_pc *pc)
*/
u32 xe_guc_pc_get_rpe_freq(struct xe_guc_pc *pc)
{
pc_update_rp_values(pc);
struct xe_device *xe = pc_to_xe(pc);
u32 freq;
return pc->rpe_freq;
if (GRAPHICS_VERx100(xe) == 1260)
freq = pvc_get_rpe_freq(pc);
else if (GRAPHICS_VERx100(xe) >= 1270)
freq = mtl_get_rpe_freq(pc);
else
freq = tgl_get_rpe_freq(pc);
return freq;
}
/**
@ -1022,7 +1024,7 @@ static int pc_set_mert_freq_cap(struct xe_guc_pc *pc)
/*
* Ensure min and max are bound by MERT_FREQ_CAP until driver loads.
*/
ret = pc_set_min_freq(pc, min(pc->rpe_freq, pc_max_freq_cap(pc)));
ret = pc_set_min_freq(pc, min(xe_guc_pc_get_rpe_freq(pc), pc_max_freq_cap(pc)));
if (!ret)
ret = pc_set_max_freq(pc, min(pc->rp0_freq, pc_max_freq_cap(pc)));
@ -1133,8 +1135,6 @@ static int pc_init_freqs(struct xe_guc_pc *pc)
if (ret)
goto out;
pc_update_rp_values(pc);
pc_init_pcode_freq(pc);
/*
@ -1340,7 +1340,7 @@ static void xe_guc_pc_fini_hw(void *arg)
XE_WARN_ON(xe_guc_pc_stop(pc));
/* Bind requested freq to mert_freq_cap before unload */
pc_set_cur_freq(pc, min(pc_max_freq_cap(pc), pc->rpe_freq));
pc_set_cur_freq(pc, min(pc_max_freq_cap(pc), xe_guc_pc_get_rpe_freq(pc)));
xe_force_wake_put(gt_to_fw(pc_to_gt(pc)), fw_ref);
}

4
drivers/gpu/drm/xe/xe_guc_pc_types.h
View File

@ -19,10 +19,6 @@ struct xe_guc_pc {
atomic_t flush_freq_limit;
/** @rp0_freq: HW RP0 frequency - The Maximum one */
u32 rp0_freq;
/** @rpa_freq: HW RPa frequency - The Achievable one */
u32 rpa_freq;
/** @rpe_freq: HW RPe frequency - The Efficient one */
u32 rpe_freq;
/** @rpn_freq: HW RPN frequency - The Minimum one */
u32 rpn_freq;
/** @user_requested_min: Stash the minimum requested freq by user */

128
drivers/gpu/drm/xe/xe_migrate.c
View File

@ -29,6 +29,7 @@
#include "xe_lrc.h"
#include "xe_map.h"
#include "xe_mocs.h"
#include "xe_printk.h"
#include "xe_pt.h"
#include "xe_res_cursor.h"
#include "xe_sa.h"
@ -1210,6 +1211,128 @@ struct xe_exec_queue *xe_migrate_exec_queue(struct xe_migrate *migrate)
return migrate->q;
}
/**
* xe_migrate_vram_copy_chunk() - Copy a chunk of a VRAM buffer object.
* @vram_bo: The VRAM buffer object.
* @vram_offset: The VRAM offset.
* @sysmem_bo: The sysmem buffer object.
* @sysmem_offset: The sysmem offset.
* @size: The size of VRAM chunk to copy.
* @dir: The direction of the copy operation.
*
* Copies a portion of a buffer object between VRAM and system memory.
* On Xe2 platforms that support flat CCS, VRAM data is decompressed when
* copying to system memory.
*
* Return: Pointer to a dma_fence representing the last copy batch, or
* an error pointer on failure. If there is a failure, any copy operation
* started by the function call has been synced.
*/
struct dma_fence *xe_migrate_vram_copy_chunk(struct xe_bo *vram_bo, u64 vram_offset,
struct xe_bo *sysmem_bo, u64 sysmem_offset,
u64 size, enum xe_migrate_copy_dir dir)
{
struct xe_device *xe = xe_bo_device(vram_bo);
struct xe_tile *tile = vram_bo->tile;
struct xe_gt *gt = tile->primary_gt;
struct xe_migrate *m = tile->migrate;
struct dma_fence *fence = NULL;
struct ttm_resource *vram = vram_bo->ttm.resource;
struct ttm_resource *sysmem = sysmem_bo->ttm.resource;
struct xe_res_cursor vram_it, sysmem_it;
u64 vram_L0_ofs, sysmem_L0_ofs;
u32 vram_L0_pt, sysmem_L0_pt;
u64 vram_L0, sysmem_L0;
bool to_sysmem = (dir == XE_MIGRATE_COPY_TO_SRAM);
bool use_comp_pat = to_sysmem &&
GRAPHICS_VER(xe) >= 20 && xe_device_has_flat_ccs(xe);
int pass = 0;
int err;
xe_assert(xe, IS_ALIGNED(vram_offset | sysmem_offset | size, PAGE_SIZE));
xe_assert(xe, xe_bo_is_vram(vram_bo));
xe_assert(xe, !xe_bo_is_vram(sysmem_bo));
xe_assert(xe, !range_overflows(vram_offset, size, (u64)vram_bo->ttm.base.size));
xe_assert(xe, !range_overflows(sysmem_offset, size, (u64)sysmem_bo->ttm.base.size));
xe_res_first(vram, vram_offset, size, &vram_it);
xe_res_first_sg(xe_bo_sg(sysmem_bo), sysmem_offset, size, &sysmem_it);
while (size) {
u32 pte_flags = PTE_UPDATE_FLAG_IS_VRAM;
u32 batch_size = 2; /* arb_clear() + MI_BATCH_BUFFER_END */
struct xe_sched_job *job;
struct xe_bb *bb;
u32 update_idx;
bool usm = xe->info.has_usm;
u32 avail_pts = max_mem_transfer_per_pass(xe) / LEVEL0_PAGE_TABLE_ENCODE_SIZE;
sysmem_L0 = xe_migrate_res_sizes(m, &sysmem_it);
vram_L0 = min(xe_migrate_res_sizes(m, &vram_it), sysmem_L0);
xe_dbg(xe, "Pass %u, size: %llu\n", pass++, vram_L0);
pte_flags |= use_comp_pat ? PTE_UPDATE_FLAG_IS_COMP_PTE : 0;
batch_size += pte_update_size(m, pte_flags, vram, &vram_it, &vram_L0,
&vram_L0_ofs, &vram_L0_pt, 0, 0, avail_pts);
batch_size += pte_update_size(m, 0, sysmem, &sysmem_it, &vram_L0, &sysmem_L0_ofs,
&sysmem_L0_pt, 0, avail_pts, avail_pts);
batch_size += EMIT_COPY_DW;
bb = xe_bb_new(gt, batch_size, usm);
if (IS_ERR(bb)) {
err = PTR_ERR(bb);
return ERR_PTR(err);
}
if (xe_migrate_allow_identity(vram_L0, &vram_it))
xe_res_next(&vram_it, vram_L0);
else
emit_pte(m, bb, vram_L0_pt, true, use_comp_pat, &vram_it, vram_L0, vram);
emit_pte(m, bb, sysmem_L0_pt, false, false, &sysmem_it, vram_L0, sysmem);
bb->cs[bb->len++] = MI_BATCH_BUFFER_END;
update_idx = bb->len;
if (to_sysmem)
emit_copy(gt, bb, vram_L0_ofs, sysmem_L0_ofs, vram_L0, XE_PAGE_SIZE);
else
emit_copy(gt, bb, sysmem_L0_ofs, vram_L0_ofs, vram_L0, XE_PAGE_SIZE);
job = xe_bb_create_migration_job(m->q, bb, xe_migrate_batch_base(m, usm),
update_idx);
if (IS_ERR(job)) {
xe_bb_free(bb, NULL);
err = PTR_ERR(job);
return ERR_PTR(err);
}
xe_sched_job_add_migrate_flush(job, MI_INVALIDATE_TLB);
xe_assert(xe, dma_resv_test_signaled(vram_bo->ttm.base.resv,
DMA_RESV_USAGE_BOOKKEEP));
xe_assert(xe, dma_resv_test_signaled(sysmem_bo->ttm.base.resv,
DMA_RESV_USAGE_BOOKKEEP));
scoped_guard(mutex, &m->job_mutex) {
xe_sched_job_arm(job);
dma_fence_put(fence);
fence = dma_fence_get(&job->drm.s_fence->finished);
xe_sched_job_push(job);
dma_fence_put(m->fence);
m->fence = dma_fence_get(fence);
}
xe_bb_free(bb, fence);
size -= vram_L0;
}
return fence;
}
static void emit_clear_link_copy(struct xe_gt *gt, struct xe_bb *bb, u64 src_ofs,
u32 size, u32 pitch)
{
@ -1912,11 +2035,6 @@ static bool xe_migrate_vram_use_pde(struct drm_pagemap_addr *sram_addr,
return true;
}
enum xe_migrate_copy_dir {
XE_MIGRATE_COPY_TO_VRAM,
XE_MIGRATE_COPY_TO_SRAM,
};
#define XE_CACHELINE_BYTES 64ull
#define XE_CACHELINE_MASK (XE_CACHELINE_BYTES - 1)

8
drivers/gpu/drm/xe/xe_migrate.h
View File

@ -28,6 +28,11 @@ struct xe_vma;
enum xe_sriov_vf_ccs_rw_ctxs;
enum xe_migrate_copy_dir {
XE_MIGRATE_COPY_TO_VRAM,
XE_MIGRATE_COPY_TO_SRAM,
};
/**
* struct xe_migrate_pt_update_ops - Callbacks for the
* xe_migrate_update_pgtables() function.
@ -131,6 +136,9 @@ int xe_migrate_ccs_rw_copy(struct xe_tile *tile, struct xe_exec_queue *q,
struct xe_lrc *xe_migrate_lrc(struct xe_migrate *migrate);
struct xe_exec_queue *xe_migrate_exec_queue(struct xe_migrate *migrate);
struct dma_fence *xe_migrate_vram_copy_chunk(struct xe_bo *vram_bo, u64 vram_offset,
struct xe_bo *sysmem_bo, u64 sysmem_offset,
u64 size, enum xe_migrate_copy_dir dir);
int xe_migrate_access_memory(struct xe_migrate *m, struct xe_bo *bo,
unsigned long offset, void *buf, int len,
int write);

9
drivers/gpu/drm/xe/xe_oa.c
View File

@ -870,7 +870,7 @@ static void xe_oa_stream_destroy(struct xe_oa_stream *stream)
xe_oa_free_oa_buffer(stream);
xe_force_wake_put(gt_to_fw(gt), XE_FORCEWAKE_ALL);
xe_force_wake_put(gt_to_fw(gt), stream->fw_ref);
xe_pm_runtime_put(stream->oa->xe);
/* Wa_1509372804:pvc: Unset the override of GUCRC mode to enable rc6 */
@ -1717,7 +1717,6 @@ static int xe_oa_stream_init(struct xe_oa_stream *stream,
struct xe_oa_open_param *param)
{
struct xe_gt *gt = param->hwe->gt;
unsigned int fw_ref;
int ret;
stream->exec_q = param->exec_q;
@ -1772,8 +1771,8 @@ static int xe_oa_stream_init(struct xe_oa_stream *stream,
/* Take runtime pm ref and forcewake to disable RC6 */
xe_pm_runtime_get(stream->oa->xe);
fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
if (!xe_force_wake_ref_has_domain(fw_ref, XE_FORCEWAKE_ALL)) {
stream->fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
if (!xe_force_wake_ref_has_domain(stream->fw_ref, XE_FORCEWAKE_ALL)) {
ret = -ETIMEDOUT;
goto err_fw_put;
}
@ -1818,7 +1817,7 @@ static int xe_oa_stream_init(struct xe_oa_stream *stream,
err_free_oa_buf:
xe_oa_free_oa_buffer(stream);
err_fw_put:
xe_force_wake_put(gt_to_fw(gt), fw_ref);
xe_force_wake_put(gt_to_fw(gt), stream->fw_ref);
xe_pm_runtime_put(stream->oa->xe);
if (stream->override_gucrc)
xe_gt_WARN_ON(gt, xe_guc_pc_unset_gucrc_mode(&gt->uc.guc.pc));

3
drivers/gpu/drm/xe/xe_oa_types.h
View File

@ -264,5 +264,8 @@ struct xe_oa_stream {
/** @syncs: syncs to wait on and to signal */
struct xe_sync_entry *syncs;
/** @fw_ref: Forcewake reference */
unsigned int fw_ref;
};
#endif

2
drivers/gpu/drm/xe/xe_pci.c
View File

@ -333,7 +333,7 @@ static const struct xe_device_desc mtl_desc = {
.has_pxp = true,
.max_gt_per_tile = 2,
.va_bits = 48,
.vm_max_level = 4,
.vm_max_level = 3,
};
static const struct xe_device_desc lnl_desc = {

21
drivers/gpu/drm/xe/xe_pci_sriov.c
View File

@ -94,6 +94,20 @@ static int resize_vf_vram_bar(struct xe_device *xe, int num_vfs)
return pci_iov_vf_bar_set_size(pdev, VF_LMEM_BAR, __fls(sizes));
}
static int pf_prepare_vfs_enabling(struct xe_device *xe)
{
xe_assert(xe, IS_SRIOV_PF(xe));
/* make sure we are not locked-down by other components */
return xe_sriov_pf_arm_guard(xe, &xe->sriov.pf.guard_vfs_enabling, false, NULL);
}
static void pf_finish_vfs_enabling(struct xe_device *xe)
{
xe_assert(xe, IS_SRIOV_PF(xe));
/* allow other components to lockdown VFs enabling */
xe_sriov_pf_disarm_guard(xe, &xe->sriov.pf.guard_vfs_enabling, false, NULL);
}
static int pf_enable_vfs(struct xe_device *xe, int num_vfs)
{
struct pci_dev *pdev = to_pci_dev(xe->drm.dev);
@ -109,6 +123,10 @@ static int pf_enable_vfs(struct xe_device *xe, int num_vfs)
if (err)
goto out;
err = pf_prepare_vfs_enabling(xe);
if (err)
goto out;
/*
* We must hold additional reference to the runtime PM to keep PF in D0
* during VFs lifetime, as our VFs do not implement the PM capability.
@ -148,6 +166,7 @@ static int pf_enable_vfs(struct xe_device *xe, int num_vfs)
failed:
xe_sriov_pf_unprovision_vfs(xe, num_vfs);
xe_pm_runtime_put(xe);
pf_finish_vfs_enabling(xe);
out:
xe_sriov_notice(xe, "Failed to enable %u VF%s (%pe)\n",
num_vfs, str_plural(num_vfs), ERR_PTR(err));
@ -179,6 +198,8 @@ static int pf_disable_vfs(struct xe_device *xe)
/* not needed anymore - see pf_enable_vfs() */
xe_pm_runtime_put(xe);
pf_finish_vfs_enabling(xe);
xe_sriov_info(xe, "Disabled %u VF%s\n", num_vfs, str_plural(num_vfs));
return 0;
}

40
drivers/gpu/drm/xe/xe_pcode.c
View File

@ -32,27 +32,39 @@
static int pcode_mailbox_status(struct xe_tile *tile)
{
const char *err_str;
int err_decode;
u32 err;
static const struct pcode_err_decode err_decode[] = {
[PCODE_ILLEGAL_CMD] = {-ENXIO, "Illegal Command"},
[PCODE_TIMEOUT] = {-ETIMEDOUT, "Timed out"},
[PCODE_ILLEGAL_DATA] = {-EINVAL, "Illegal Data"},
[PCODE_ILLEGAL_SUBCOMMAND] = {-ENXIO, "Illegal Subcommand"},
[PCODE_LOCKED] = {-EBUSY, "PCODE Locked"},
[PCODE_GT_RATIO_OUT_OF_RANGE] = {-EOVERFLOW,
"GT ratio out of range"},
[PCODE_REJECTED] = {-EACCES, "PCODE Rejected"},
[PCODE_ERROR_MASK] = {-EPROTO, "Unknown"},
};
#define CASE_ERR(_err, _err_decode, _err_str) \
case _err: \
err_decode = _err_decode; \
err_str = _err_str; \
break
err = xe_mmio_read32(&tile->mmio, PCODE_MAILBOX) & PCODE_ERROR_MASK;
switch (err) {
CASE_ERR(PCODE_ILLEGAL_CMD, -ENXIO, "Illegal Command");
CASE_ERR(PCODE_TIMEOUT, -ETIMEDOUT, "Timed out");
CASE_ERR(PCODE_ILLEGAL_DATA, -EINVAL, "Illegal Data");
CASE_ERR(PCODE_ILLEGAL_SUBCOMMAND, -ENXIO, "Illegal Subcommand");
CASE_ERR(PCODE_LOCKED, -EBUSY, "PCODE Locked");
CASE_ERR(PCODE_GT_RATIO_OUT_OF_RANGE, -EOVERFLOW, "GT ratio out of range");
CASE_ERR(PCODE_REJECTED, -EACCES, "PCODE Rejected");
default:
err_decode = -EPROTO;
err_str = "Unknown";
}
if (err) {
drm_err(&tile_to_xe(tile)->drm, "PCODE Mailbox failed: %d %s", err,
err_decode[err].str ?: "Unknown");
return err_decode[err].errno ?: -EPROTO;
drm_err(&tile_to_xe(tile)->drm, "PCODE Mailbox failed: %d %s",
err_decode, err_str);
return err_decode;
}
return 0;
#undef CASE_ERR
}
static int __pcode_mailbox_rw(struct xe_tile *tile, u32 mbox, u32 *data0, u32 *data1,

6
drivers/gpu/drm/xe/xe_pcode_api.h
View File

@ -92,9 +92,3 @@
#define BMG_PCIE_CAP XE_REG(0x138340)
#define LINK_DOWNGRADE REG_GENMASK(1, 0)
#define DOWNGRADE_CAPABLE 2
struct pcode_err_decode {
int errno;
const char *str;
};

2
drivers/gpu/drm/xe/xe_pm.c
View File

@ -112,7 +112,7 @@ void xe_pm_might_block_on_suspend(void)
}
/**
* xe_pm_might_block_on_suspend() - Block pending suspend.
* xe_pm_block_on_suspend() - Block pending suspend.
* @xe: The xe device about to be suspended.
*
* Block if the pm notifier has start evicting bos, to avoid

21
drivers/gpu/drm/xe/xe_sa.c
View File

@ -110,6 +110,10 @@ struct drm_suballoc *__xe_sa_bo_new(struct xe_sa_manager *sa_manager, u32 size,
return drm_suballoc_new(&sa_manager->base, size, gfp, true, 0);
}
/**
* xe_sa_bo_flush_write() - Copy the data from the sub-allocation to the GPU memory.
* @sa_bo: the &drm_suballoc to flush
*/
void xe_sa_bo_flush_write(struct drm_suballoc *sa_bo)
{
struct xe_sa_manager *sa_manager = to_xe_sa_manager(sa_bo->manager);
@ -123,6 +127,23 @@ void xe_sa_bo_flush_write(struct drm_suballoc *sa_bo)
drm_suballoc_size(sa_bo));
}
/**
* xe_sa_bo_sync_read() - Copy the data from GPU memory to the sub-allocation.
* @sa_bo: the &drm_suballoc to sync
*/
void xe_sa_bo_sync_read(struct drm_suballoc *sa_bo)
{
struct xe_sa_manager *sa_manager = to_xe_sa_manager(sa_bo->manager);
struct xe_device *xe = tile_to_xe(sa_manager->bo->tile);
if (!sa_manager->bo->vmap.is_iomem)
return;
xe_map_memcpy_from(xe, xe_sa_bo_cpu_addr(sa_bo), &sa_manager->bo->vmap,
drm_suballoc_soffset(sa_bo),
drm_suballoc_size(sa_bo));
}
void xe_sa_bo_free(struct drm_suballoc *sa_bo,
struct dma_fence *fence)
{

1
drivers/gpu/drm/xe/xe_sa.h
View File

@ -37,6 +37,7 @@ static inline struct drm_suballoc *xe_sa_bo_new(struct xe_sa_manager *sa_manager
}
void xe_sa_bo_flush_write(struct drm_suballoc *sa_bo);
void xe_sa_bo_sync_read(struct drm_suballoc *sa_bo);
void xe_sa_bo_free(struct drm_suballoc *sa_bo, struct dma_fence *fence);
static inline struct xe_sa_manager *

520
drivers/gpu/drm/xe/xe_sriov_packet.c Normal file
View File

@ -0,0 +1,520 @@
// SPDX-License-Identifier: MIT
/*
* Copyright © 2025 Intel Corporation
*/
#include "xe_bo.h"
#include "xe_device.h"
#include "xe_guc_klv_helpers.h"
#include "xe_printk.h"
#include "xe_sriov_packet.h"
#include "xe_sriov_packet_types.h"
#include "xe_sriov_pf_helpers.h"
#include "xe_sriov_pf_migration.h"
#include "xe_sriov_printk.h"
static struct mutex *pf_migration_mutex(struct xe_device *xe, unsigned int vfid)
{
xe_assert(xe, IS_SRIOV_PF(xe));
xe_assert(xe, vfid <= xe_sriov_pf_get_totalvfs(xe));
return &xe->sriov.pf.vfs[vfid].migration.lock;
}
static struct xe_sriov_packet **pf_pick_pending(struct xe_device *xe, unsigned int vfid)
{
xe_assert(xe, IS_SRIOV_PF(xe));
xe_assert(xe, vfid <= xe_sriov_pf_get_totalvfs(xe));
lockdep_assert_held(pf_migration_mutex(xe, vfid));
return &xe->sriov.pf.vfs[vfid].migration.pending;
}
static struct xe_sriov_packet **
pf_pick_descriptor(struct xe_device *xe, unsigned int vfid)
{
xe_assert(xe, IS_SRIOV_PF(xe));
xe_assert(xe, vfid <= xe_sriov_pf_get_totalvfs(xe));
lockdep_assert_held(pf_migration_mutex(xe, vfid));
return &xe->sriov.pf.vfs[vfid].migration.descriptor;
}
static struct xe_sriov_packet **pf_pick_trailer(struct xe_device *xe, unsigned int vfid)
{
xe_assert(xe, IS_SRIOV_PF(xe));
xe_assert(xe, vfid <= xe_sriov_pf_get_totalvfs(xe));
lockdep_assert_held(pf_migration_mutex(xe, vfid));
return &xe->sriov.pf.vfs[vfid].migration.trailer;
}
static struct xe_sriov_packet **pf_pick_read_packet(struct xe_device *xe,
unsigned int vfid)
{
struct xe_sriov_packet **data;
data = pf_pick_descriptor(xe, vfid);
if (*data)
return data;
data = pf_pick_pending(xe, vfid);
if (!*data)
*data = xe_sriov_pf_migration_save_consume(xe, vfid);
if (*data)
return data;
data = pf_pick_trailer(xe, vfid);
if (*data)
return data;
return NULL;
}
static bool pkt_needs_bo(struct xe_sriov_packet *data)
{
return data->hdr.type == XE_SRIOV_PACKET_TYPE_VRAM;
}
/**
* xe_sriov_packet_alloc() - Allocate migration data packet
* @xe: the &xe_device
*
* Only allocates the "outer" structure, without initializing the migration
* data backing storage.
*
* Return: Pointer to &xe_sriov_packet on success,
* NULL in case of error.
*/
struct xe_sriov_packet *xe_sriov_packet_alloc(struct xe_device *xe)
{
struct xe_sriov_packet *data;
data = kzalloc(sizeof(*data), GFP_KERNEL);
if (!data)
return NULL;
data->xe = xe;
data->hdr_remaining = sizeof(data->hdr);
return data;
}
/**
* xe_sriov_packet_free() - Free migration data packet.
* @data: the &xe_sriov_packet
*/
void xe_sriov_packet_free(struct xe_sriov_packet *data)
{
if (IS_ERR_OR_NULL(data))
return;
if (pkt_needs_bo(data))
xe_bo_unpin_map_no_vm(data->bo);
else
kvfree(data->buff);
kfree(data);
}
static int pkt_init(struct xe_sriov_packet *data)
{
struct xe_gt *gt = xe_device_get_gt(data->xe, data->hdr.gt_id);
if (!gt)
return -EINVAL;
if (data->hdr.size == 0)
return 0;
if (pkt_needs_bo(data)) {
struct xe_bo *bo;
bo = xe_bo_create_pin_map_novm(data->xe, gt->tile, PAGE_ALIGN(data->hdr.size),
ttm_bo_type_kernel,
XE_BO_FLAG_SYSTEM | XE_BO_FLAG_PINNED, false);
if (IS_ERR(bo))
return PTR_ERR(bo);
data->bo = bo;
data->vaddr = bo->vmap.vaddr;
} else {
void *buff = kvzalloc(data->hdr.size, GFP_KERNEL);
if (!buff)
return -ENOMEM;
data->buff = buff;
data->vaddr = buff;
}
return 0;
}
#define XE_SRIOV_PACKET_SUPPORTED_VERSION 1
/**
* xe_sriov_packet_init() - Initialize migration packet header and backing storage.
* @data: the &xe_sriov_packet
* @tile_id: tile identifier
* @gt_id: GT identifier
* @type: &xe_sriov_packet_type
* @offset: offset of data packet payload (within wider resource)
* @size: size of data packet payload
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_packet_init(struct xe_sriov_packet *data, u8 tile_id, u8 gt_id,
enum xe_sriov_packet_type type, loff_t offset, size_t size)
{
data->hdr.version = XE_SRIOV_PACKET_SUPPORTED_VERSION;
data->hdr.type = type;
data->hdr.tile_id = tile_id;
data->hdr.gt_id = gt_id;
data->hdr.offset = offset;
data->hdr.size = size;
data->remaining = size;
return pkt_init(data);
}
/**
* xe_sriov_packet_init_from_hdr() - Initialize migration packet backing storage based on header.
* @data: the &xe_sriov_packet
*
* Header data is expected to be filled prior to calling this function.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_packet_init_from_hdr(struct xe_sriov_packet *data)
{
xe_assert(data->xe, !data->hdr_remaining);
if (data->hdr.version != XE_SRIOV_PACKET_SUPPORTED_VERSION)
return -EINVAL;
data->remaining = data->hdr.size;
return pkt_init(data);
}
static ssize_t pkt_hdr_read(struct xe_sriov_packet *data,
char __user *buf, size_t len)
{
loff_t offset = sizeof(data->hdr) - data->hdr_remaining;
if (!data->hdr_remaining)
return -EINVAL;
if (len > data->hdr_remaining)
len = data->hdr_remaining;
if (copy_to_user(buf, (void *)&data->hdr + offset, len))
return -EFAULT;
data->hdr_remaining -= len;
return len;
}
static ssize_t pkt_data_read(struct xe_sriov_packet *data,
char __user *buf, size_t len)
{
if (len > data->remaining)
len = data->remaining;
if (copy_to_user(buf, data->vaddr + (data->hdr.size - data->remaining), len))
return -EFAULT;
data->remaining -= len;
return len;
}
static ssize_t pkt_read_single(struct xe_sriov_packet **data,
unsigned int vfid, char __user *buf, size_t len)
{
ssize_t copied = 0;
if ((*data)->hdr_remaining)
copied = pkt_hdr_read(*data, buf, len);
else
copied = pkt_data_read(*data, buf, len);
if ((*data)->remaining == 0 && (*data)->hdr_remaining == 0) {
xe_sriov_packet_free(*data);
*data = NULL;
}
return copied;
}
/**
* xe_sriov_packet_read_single() - Read migration data from a single packet.
* @xe: the &xe_device
* @vfid: the VF identifier
* @buf: start address of userspace buffer
* @len: requested read size from userspace
*
* Return: number of bytes that has been successfully read,
* 0 if no more migration data is available,
* -errno on failure.
*/
ssize_t xe_sriov_packet_read_single(struct xe_device *xe, unsigned int vfid,
char __user *buf, size_t len)
{
struct xe_sriov_packet **data = pf_pick_read_packet(xe, vfid);
if (!data)
return -ENODATA;
if (IS_ERR(*data))
return PTR_ERR(*data);
return pkt_read_single(data, vfid, buf, len);
}
static ssize_t pkt_hdr_write(struct xe_sriov_packet *data,
const char __user *buf, size_t len)
{
loff_t offset = sizeof(data->hdr) - data->hdr_remaining;
int ret;
if (len > data->hdr_remaining)
len = data->hdr_remaining;
if (copy_from_user((void *)&data->hdr + offset, buf, len))
return -EFAULT;
data->hdr_remaining -= len;
if (!data->hdr_remaining) {
ret = xe_sriov_packet_init_from_hdr(data);
if (ret)
return ret;
}
return len;
}
static ssize_t pkt_data_write(struct xe_sriov_packet *data,
const char __user *buf, size_t len)
{
if (len > data->remaining)
len = data->remaining;
if (copy_from_user(data->vaddr + (data->hdr.size - data->remaining), buf, len))
return -EFAULT;
data->remaining -= len;
return len;
}
/**
* xe_sriov_packet_write_single() - Write migration data to a single packet.
* @xe: the &xe_device
* @vfid: the VF identifier
* @buf: start address of userspace buffer
* @len: requested write size from userspace
*
* Return: number of bytes that has been successfully written,
* -errno on failure.
*/
ssize_t xe_sriov_packet_write_single(struct xe_device *xe, unsigned int vfid,
const char __user *buf, size_t len)
{
struct xe_sriov_packet **data = pf_pick_pending(xe, vfid);
int ret;
ssize_t copied;
if (IS_ERR_OR_NULL(*data)) {
*data = xe_sriov_packet_alloc(xe);
if (!*data)
return -ENOMEM;
}
if ((*data)->hdr_remaining)
copied = pkt_hdr_write(*data, buf, len);
else
copied = pkt_data_write(*data, buf, len);
if ((*data)->hdr_remaining == 0 && (*data)->remaining == 0) {
ret = xe_sriov_pf_migration_restore_produce(xe, vfid, *data);
if (ret) {
xe_sriov_packet_free(*data);
return ret;
}
*data = NULL;
}
return copied;
}
#define MIGRATION_KLV_DEVICE_DEVID_KEY 0xf001u
#define MIGRATION_KLV_DEVICE_DEVID_LEN 1u
#define MIGRATION_KLV_DEVICE_REVID_KEY 0xf002u
#define MIGRATION_KLV_DEVICE_REVID_LEN 1u
#define MIGRATION_DESCRIPTOR_DWORDS (GUC_KLV_LEN_MIN + MIGRATION_KLV_DEVICE_DEVID_LEN + \
GUC_KLV_LEN_MIN + MIGRATION_KLV_DEVICE_REVID_LEN)
static size_t pf_descriptor_init(struct xe_device *xe, unsigned int vfid)
{
struct xe_sriov_packet **desc = pf_pick_descriptor(xe, vfid);
struct xe_sriov_packet *data;
unsigned int len = 0;
u32 *klvs;
int ret;
data = xe_sriov_packet_alloc(xe);
if (!data)
return -ENOMEM;
ret = xe_sriov_packet_init(data, 0, 0, XE_SRIOV_PACKET_TYPE_DESCRIPTOR,
0, MIGRATION_DESCRIPTOR_DWORDS * sizeof(u32));
if (ret) {
xe_sriov_packet_free(data);
return ret;
}
klvs = data->vaddr;
klvs[len++] = PREP_GUC_KLV_CONST(MIGRATION_KLV_DEVICE_DEVID_KEY,
MIGRATION_KLV_DEVICE_DEVID_LEN);
klvs[len++] = xe->info.devid;
klvs[len++] = PREP_GUC_KLV_CONST(MIGRATION_KLV_DEVICE_REVID_KEY,
MIGRATION_KLV_DEVICE_REVID_LEN);
klvs[len++] = xe->info.revid;
xe_assert(xe, len == MIGRATION_DESCRIPTOR_DWORDS);
*desc = data;
return 0;
}
/**
* xe_sriov_packet_process_descriptor() - Process migration data descriptor packet.
* @xe: the &xe_device
* @vfid: the VF identifier
* @data: the &xe_sriov_packet containing the descriptor
*
* The descriptor uses the same KLV format as GuC, and contains metadata used for
* checking migration data compatibility.
*
* Return: 0 on success, -errno on failure.
*/
int xe_sriov_packet_process_descriptor(struct xe_device *xe, unsigned int vfid,
struct xe_sriov_packet *data)
{
u32 num_dwords = data->hdr.size / sizeof(u32);
u32 *klvs = data->vaddr;
xe_assert(xe, data->hdr.type == XE_SRIOV_PACKET_TYPE_DESCRIPTOR);
if (data->hdr.size % sizeof(u32)) {
xe_sriov_warn(xe, "Aborting migration, descriptor not in KLV format (size=%llu)\n",
data->hdr.size);
return -EINVAL;
}
while (num_dwords >= GUC_KLV_LEN_MIN) {
u32 key = FIELD_GET(GUC_KLV_0_KEY, klvs[0]);
u32 len = FIELD_GET(GUC_KLV_0_LEN, klvs[0]);
klvs += GUC_KLV_LEN_MIN;
num_dwords -= GUC_KLV_LEN_MIN;
if (len > num_dwords) {
xe_sriov_warn(xe, "Aborting migration, truncated KLV %#x, len %u\n",
key, len);
return -EINVAL;
}
switch (key) {
case MIGRATION_KLV_DEVICE_DEVID_KEY:
if (*klvs != xe->info.devid) {
xe_sriov_warn(xe,
"Aborting migration, devid mismatch %#06x!=%#06x\n",
*klvs, xe->info.devid);
return -ENODEV;
}
break;
case MIGRATION_KLV_DEVICE_REVID_KEY:
if (*klvs != xe->info.revid) {
xe_sriov_warn(xe,
"Aborting migration, revid mismatch %#06x!=%#06x\n",
*klvs, xe->info.revid);
return -ENODEV;
}
break;
default:
xe_sriov_dbg(xe,
"Skipping unknown migration KLV %#x, len=%u\n",
key, len);
print_hex_dump_bytes("desc: ", DUMP_PREFIX_OFFSET, klvs,
min(SZ_64, len * sizeof(u32)));
break;
}
klvs += len;
num_dwords -= len;
}
return 0;
}
static void pf_pending_init(struct xe_device *xe, unsigned int vfid)
{
struct xe_sriov_packet **data = pf_pick_pending(xe, vfid);
*data = NULL;
}
#define MIGRATION_TRAILER_SIZE 0
static int pf_trailer_init(struct xe_device *xe, unsigned int vfid)
{
struct xe_sriov_packet **trailer = pf_pick_trailer(xe, vfid);
struct xe_sriov_packet *data;
int ret;
data = xe_sriov_packet_alloc(xe);
if (!data)
return -ENOMEM;
ret = xe_sriov_packet_init(data, 0, 0, XE_SRIOV_PACKET_TYPE_TRAILER,
0, MIGRATION_TRAILER_SIZE);
if (ret) {
xe_sriov_packet_free(data);
return ret;
}
*trailer = data;
return 0;
}
/**
* xe_sriov_packet_save_init() - Initialize the pending save migration packets.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* Return: 0 on success, -errno on failure.
*/
int xe_sriov_packet_save_init(struct xe_device *xe, unsigned int vfid)
{
int ret;
scoped_cond_guard(mutex_intr, return -EINTR, pf_migration_mutex(xe, vfid)) {
ret = pf_descriptor_init(xe, vfid);
if (ret)
return ret;
ret = pf_trailer_init(xe, vfid);
if (ret)
return ret;
pf_pending_init(xe, vfid);
}
return 0;
}

30
drivers/gpu/drm/xe/xe_sriov_packet.h Normal file
View File

@ -0,0 +1,30 @@
/* SPDX-License-Identifier: MIT */
/*
* Copyright © 2025 Intel Corporation
*/
#ifndef _XE_SRIOV_PACKET_H_
#define _XE_SRIOV_PACKET_H_
#include <linux/types.h>
struct xe_device;
struct xe_sriov_packet;
enum xe_sriov_packet_type;
struct xe_sriov_packet *xe_sriov_packet_alloc(struct xe_device *xe);
void xe_sriov_packet_free(struct xe_sriov_packet *data);
int xe_sriov_packet_init(struct xe_sriov_packet *data, u8 tile_id, u8 gt_id,
enum xe_sriov_packet_type, loff_t offset, size_t size);
int xe_sriov_packet_init_from_hdr(struct xe_sriov_packet *data);
ssize_t xe_sriov_packet_read_single(struct xe_device *xe, unsigned int vfid,
char __user *buf, size_t len);
ssize_t xe_sriov_packet_write_single(struct xe_device *xe, unsigned int vfid,
const char __user *buf, size_t len);
int xe_sriov_packet_save_init(struct xe_device *xe, unsigned int vfid);
int xe_sriov_packet_process_descriptor(struct xe_device *xe, unsigned int vfid,
struct xe_sriov_packet *data);
#endif

75
drivers/gpu/drm/xe/xe_sriov_packet_types.h Normal file
View File

@ -0,0 +1,75 @@
/* SPDX-License-Identifier: MIT */
/*
* Copyright © 2025 Intel Corporation
*/
#ifndef _XE_SRIOV_PACKET_TYPES_H_
#define _XE_SRIOV_PACKET_TYPES_H_
#include <linux/types.h>
/**
* enum xe_sriov_packet_type - Xe SR-IOV VF migration data packet type
* @XE_SRIOV_PACKET_TYPE_DESCRIPTOR: Descriptor with VF device metadata
* @XE_SRIOV_PACKET_TYPE_TRAILER: Trailer indicating end-of-stream
* @XE_SRIOV_PACKET_TYPE_GGTT: Global GTT migration data
* @XE_SRIOV_PACKET_TYPE_MMIO: MMIO registers migration data
* @XE_SRIOV_PACKET_TYPE_GUC: GuC firmware migration data
* @XE_SRIOV_PACKET_TYPE_VRAM: VRAM migration data
*/
enum xe_sriov_packet_type {
/* Skipping 0 to catch uninitialized data */
XE_SRIOV_PACKET_TYPE_DESCRIPTOR = 1,
XE_SRIOV_PACKET_TYPE_TRAILER,
XE_SRIOV_PACKET_TYPE_GGTT,
XE_SRIOV_PACKET_TYPE_MMIO,
XE_SRIOV_PACKET_TYPE_GUC,
XE_SRIOV_PACKET_TYPE_VRAM,
};
/**
* struct xe_sriov_packet_hdr - Xe SR-IOV VF migration data packet header
*/
struct xe_sriov_packet_hdr {
/** @version: migration data protocol version */
u8 version;
/** @type: migration data type */
u8 type;
/** @tile_id: migration data tile id */
u8 tile_id;
/** @gt_id: migration data gt id */
u8 gt_id;
/** @flags: migration data flags */
u32 flags;
/**
* @offset: offset into the resource;
* used when multiple packets of given type are used for migration
*/
u64 offset;
/** @size: migration data size */
u64 size;
} __packed;
/**
* struct xe_sriov_packet - Xe SR-IOV VF migration data packet
*/
struct xe_sriov_packet {
/** @xe: the PF &xe_device this data packet belongs to */
struct xe_device *xe;
/** @vaddr: CPU pointer to payload data */
void *vaddr;
/** @remaining: payload data remaining */
size_t remaining;
/** @hdr_remaining: header data remaining */
size_t hdr_remaining;
union {
/** @bo: Buffer object with migration data */
struct xe_bo *bo;
/** @buff: Buffer with migration data */
void *buff;
};
/** @hdr: data packet header */
struct xe_sriov_packet_hdr hdr;
};
#endif

102
drivers/gpu/drm/xe/xe_sriov_pf.c
View File

@ -15,6 +15,7 @@
#include "xe_sriov.h"
#include "xe_sriov_pf.h"
#include "xe_sriov_pf_helpers.h"
#include "xe_sriov_pf_migration.h"
#include "xe_sriov_pf_service.h"
#include "xe_sriov_pf_sysfs.h"
#include "xe_sriov_printk.h"
@ -102,6 +103,12 @@ int xe_sriov_pf_init_early(struct xe_device *xe)
if (err)
return err;
err = xe_sriov_pf_migration_init(xe);
if (err)
return err;
xe_guard_init(&xe->sriov.pf.guard_vfs_enabling, "vfs_enabling");
xe_sriov_pf_service_init(xe);
return 0;
@ -162,6 +169,101 @@ int xe_sriov_pf_wait_ready(struct xe_device *xe)
return 0;
}
/**
* xe_sriov_pf_arm_guard() - Arm the guard for exclusive/lockdown mode.
* @xe: the PF &xe_device
* @guard: the &xe_guard to arm
* @lockdown: arm for lockdown(true) or exclusive(false) mode
* @who: the address of the new owner, or NULL if it's a caller
*
* This function can only be called on PF.
*
* It is a simple wrapper for xe_guard_arm() with additional debug
* messages.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_arm_guard(struct xe_device *xe, struct xe_guard *guard,
bool lockdown, void *who)
{
void *new_owner = who ?: __builtin_return_address(0);
int err;
err = xe_guard_arm(guard, lockdown, new_owner);
if (err) {
xe_sriov_dbg(xe, "%s/%s mode denied (%pe) last owner %ps\n",
guard->name, xe_guard_mode_str(lockdown),
ERR_PTR(err), guard->owner);
return err;
}
xe_sriov_dbg_verbose(xe, "%s/%s by %ps\n",
guard->name, xe_guard_mode_str(lockdown),
new_owner);
return 0;
}
/**
* xe_sriov_pf_disarm_guard() - Disarm the guard.
* @xe: the PF &xe_device
* @guard: the &xe_guard to disarm
* @lockdown: disarm from lockdown(true) or exclusive(false) mode
* @who: the address of the indirect owner, or NULL if it's a caller
*
* This function can only be called on PF.
*
* It is a simple wrapper for xe_guard_disarm() with additional debug
* messages and xe_assert() to easily catch any illegal calls.
*/
void xe_sriov_pf_disarm_guard(struct xe_device *xe, struct xe_guard *guard,
bool lockdown, void *who)
{
bool disarmed;
xe_sriov_dbg_verbose(xe, "%s/%s by %ps\n",
guard->name, xe_guard_mode_str(lockdown),
who ?: __builtin_return_address(0));
disarmed = xe_guard_disarm(guard, lockdown);
xe_assert_msg(xe, disarmed, "%s/%s not armed? last owner %ps",
guard->name, xe_guard_mode_str(lockdown), guard->owner);
}
/**
* xe_sriov_pf_lockdown() - Lockdown the PF to prevent VFs enabling.
* @xe: the PF &xe_device
*
* This function can only be called on PF.
*
* Once the PF is locked down, it will not enable VFs.
* If VFs are already enabled, the -EBUSY will be returned.
* To allow the PF enable VFs again call xe_sriov_pf_end_lockdown().
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_lockdown(struct xe_device *xe)
{
xe_assert(xe, IS_SRIOV_PF(xe));
return xe_sriov_pf_arm_guard(xe, &xe->sriov.pf.guard_vfs_enabling, true,
__builtin_return_address(0));
}
/**
* xe_sriov_pf_end_lockdown() - Allow the PF to enable VFs again.
* @xe: the PF &xe_device
*
* This function can only be called on PF.
* See xe_sriov_pf_lockdown() for details.
*/
void xe_sriov_pf_end_lockdown(struct xe_device *xe)
{
xe_assert(xe, IS_SRIOV_PF(xe));
xe_sriov_pf_disarm_guard(xe, &xe->sriov.pf.guard_vfs_enabling, true,
__builtin_return_address(0));
}
/**
* xe_sriov_pf_print_vfs_summary - Print SR-IOV PF information.
* @xe: the &xe_device to print info from

4
drivers/gpu/drm/xe/xe_sriov_pf.h
View File

@ -17,11 +17,15 @@ bool xe_sriov_pf_readiness(struct xe_device *xe);
int xe_sriov_pf_init_early(struct xe_device *xe);
int xe_sriov_pf_init_late(struct xe_device *xe);
int xe_sriov_pf_wait_ready(struct xe_device *xe);
int xe_sriov_pf_lockdown(struct xe_device *xe);
void xe_sriov_pf_end_lockdown(struct xe_device *xe);
void xe_sriov_pf_print_vfs_summary(struct xe_device *xe, struct drm_printer *p);
#else
static inline bool xe_sriov_pf_readiness(struct xe_device *xe) { return false; }
static inline int xe_sriov_pf_init_early(struct xe_device *xe) { return 0; }
static inline int xe_sriov_pf_init_late(struct xe_device *xe) { return 0; }
static inline int xe_sriov_pf_lockdown(struct xe_device *xe) { return 0; }
static inline void xe_sriov_pf_end_lockdown(struct xe_device *xe) { }
#endif
#endif

128
drivers/gpu/drm/xe/xe_sriov_pf_control.c
View File

@ -5,6 +5,8 @@
#include "xe_device.h"
#include "xe_gt_sriov_pf_control.h"
#include "xe_gt_sriov_pf_migration.h"
#include "xe_sriov_packet.h"
#include "xe_sriov_pf_control.h"
#include "xe_sriov_printk.h"
@ -121,6 +123,30 @@ int xe_sriov_pf_control_reset_vf(struct xe_device *xe, unsigned int vfid)
return result;
}
/**
* xe_sriov_pf_control_wait_flr() - Wait for a VF reset (FLR) to complete.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_control_wait_flr(struct xe_device *xe, unsigned int vfid)
{
struct xe_gt *gt;
unsigned int id;
int result = 0;
int err;
for_each_gt(gt, xe, id) {
err = xe_gt_sriov_pf_control_wait_flr(gt, vfid);
result = result ? -EUCLEAN : err;
}
return result;
}
/**
* xe_sriov_pf_control_sync_flr() - Synchronize a VF FLR between all GTs.
* @xe: the &xe_device
@ -149,3 +175,105 @@ int xe_sriov_pf_control_sync_flr(struct xe_device *xe, unsigned int vfid)
return 0;
}
/**
* xe_sriov_pf_control_trigger_save_vf() - Start VF migration data SAVE sequence on all GTs.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_control_trigger_save_vf(struct xe_device *xe, unsigned int vfid)
{
struct xe_gt *gt;
unsigned int id;
int ret;
ret = xe_sriov_packet_save_init(xe, vfid);
if (ret)
return ret;
for_each_gt(gt, xe, id) {
xe_gt_sriov_pf_migration_save_init(gt, vfid);
ret = xe_gt_sriov_pf_control_trigger_save_vf(gt, vfid);
if (ret)
return ret;
}
return 0;
}
/**
* xe_sriov_pf_control_finish_save_vf() - Complete VF migration data SAVE sequence on all GTs.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_control_finish_save_vf(struct xe_device *xe, unsigned int vfid)
{
struct xe_gt *gt;
unsigned int id;
int ret;
for_each_gt(gt, xe, id) {
ret = xe_gt_sriov_pf_control_finish_save_vf(gt, vfid);
if (ret)
break;
}
return ret;
}
/**
* xe_sriov_pf_control_trigger_restore_vf() - Start VF migration data RESTORE sequence on all GTs.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_control_trigger_restore_vf(struct xe_device *xe, unsigned int vfid)
{
struct xe_gt *gt;
unsigned int id;
int ret;
for_each_gt(gt, xe, id) {
ret = xe_gt_sriov_pf_control_trigger_restore_vf(gt, vfid);
if (ret)
return ret;
}
return ret;
}
/**
* xe_sriov_pf_control_finish_restore_vf() - Complete VF migration data RESTORE sequence on all GTs.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* This function is for PF only.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_control_finish_restore_vf(struct xe_device *xe, unsigned int vfid)
{
struct xe_gt *gt;
unsigned int id;
int ret;
for_each_gt(gt, xe, id) {
ret = xe_gt_sriov_pf_control_finish_restore_vf(gt, vfid);
if (ret)
break;
}
return ret;
}

5
drivers/gpu/drm/xe/xe_sriov_pf_control.h
View File

@ -12,6 +12,11 @@ int xe_sriov_pf_control_pause_vf(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_resume_vf(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_stop_vf(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_reset_vf(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_wait_flr(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_sync_flr(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_trigger_save_vf(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_finish_save_vf(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_trigger_restore_vf(struct xe_device *xe, unsigned int vfid);
int xe_sriov_pf_control_finish_restore_vf(struct xe_device *xe, unsigned int vfid);
#endif

131
drivers/gpu/drm/xe/xe_sriov_pf_debugfs.c
View File

@ -13,6 +13,7 @@
#include "xe_sriov_pf_control.h"
#include "xe_sriov_pf_debugfs.h"
#include "xe_sriov_pf_helpers.h"
#include "xe_sriov_pf_migration.h"
#include "xe_sriov_pf_provision.h"
#include "xe_sriov_pf_service.h"
#include "xe_sriov_printk.h"
@ -98,10 +99,40 @@ static inline int xe_sriov_pf_restore_auto_provisioning(struct xe_device *xe)
DEFINE_SRIOV_ATTRIBUTE(restore_auto_provisioning);
static int lockdown_vfs_enabling_open(struct inode *inode, struct file *file)
{
struct dentry *dent = file_dentry(file);
struct xe_device *xe = extract_xe(dent);
ssize_t ret;
ret = xe_sriov_pf_lockdown(xe);
if (ret < 0)
return ret;
file->private_data = xe;
return nonseekable_open(inode, file);
}
static int lockdown_vfs_enabling_release(struct inode *inode, struct file *file)
{
struct xe_device *xe = file->private_data;
xe_sriov_pf_end_lockdown(xe);
return 0;
}
static const struct file_operations lockdown_vfs_enabling_fops = {
.owner = THIS_MODULE,
.open = lockdown_vfs_enabling_open,
.release = lockdown_vfs_enabling_release,
};
static void pf_populate_root(struct xe_device *xe, struct dentry *dent)
{
debugfs_create_file("restore_auto_provisioning", 0200, dent, xe,
&restore_auto_provisioning_fops);
debugfs_create_file("lockdown_vfs_enabling", 0400, dent, xe,
&lockdown_vfs_enabling_fops);
}
static int simple_show(struct seq_file *m, void *data)
@ -132,15 +163,36 @@ static void pf_populate_pf(struct xe_device *xe, struct dentry *pfdent)
* /sys/kernel/debug/dri/BDF/
* ├── sriov
* │ ├── vf1
* │ │ ├── migration_data
* │ │ ├── pause
* │ │ ├── reset
* │ │ ├── resume
* │ │ ├── stop
* │ │ ├── save
* │ │ ├── restore
* │ │ :
* │ ├── vf2
* │ │ ├── ...
*/
static int from_file_read_to_vf_call(struct seq_file *s,
int (*call)(struct xe_device *, unsigned int))
{
struct dentry *dent = file_dentry(s->file)->d_parent;
struct xe_device *xe = extract_xe(dent);
unsigned int vfid = extract_vfid(dent);
int ret;
xe_pm_runtime_get(xe);
ret = call(xe, vfid);
xe_pm_runtime_put(xe);
if (ret < 0)
return ret;
return 0;
}
static ssize_t from_file_write_to_vf_call(struct file *file, const char __user *userbuf,
size_t count, loff_t *ppos,
int (*call)(struct xe_device *, unsigned int))
@ -179,10 +231,85 @@ static ssize_t OP##_write(struct file *file, const char __user *userbuf, \
} \
DEFINE_SHOW_STORE_ATTRIBUTE(OP)
#define DEFINE_VF_CONTROL_ATTRIBUTE_RW(OP) \
static int OP##_show(struct seq_file *s, void *unused) \
{ \
return from_file_read_to_vf_call(s, \
xe_sriov_pf_control_finish_##OP); \
} \
static ssize_t OP##_write(struct file *file, const char __user *userbuf, \
size_t count, loff_t *ppos) \
{ \
return from_file_write_to_vf_call(file, userbuf, count, ppos, \
xe_sriov_pf_control_trigger_##OP); \
} \
DEFINE_SHOW_STORE_ATTRIBUTE(OP)
DEFINE_VF_CONTROL_ATTRIBUTE(pause_vf);
DEFINE_VF_CONTROL_ATTRIBUTE(resume_vf);
DEFINE_VF_CONTROL_ATTRIBUTE(stop_vf);
DEFINE_VF_CONTROL_ATTRIBUTE(reset_vf);
DEFINE_VF_CONTROL_ATTRIBUTE_RW(save_vf);
DEFINE_VF_CONTROL_ATTRIBUTE_RW(restore_vf);
static ssize_t data_write(struct file *file, const char __user *buf, size_t count, loff_t *pos)
{
struct dentry *dent = file_dentry(file)->d_parent;
struct xe_device *xe = extract_xe(dent);
unsigned int vfid = extract_vfid(dent);
if (*pos)
return -ESPIPE;
return xe_sriov_pf_migration_write(xe, vfid, buf, count);
}
static ssize_t data_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
{
struct dentry *dent = file_dentry(file)->d_parent;
struct xe_device *xe = extract_xe(dent);
unsigned int vfid = extract_vfid(dent);
if (*ppos)
return -ESPIPE;
return xe_sriov_pf_migration_read(xe, vfid, buf, count);
}
static const struct file_operations data_vf_fops = {
.owner = THIS_MODULE,
.open = simple_open,
.write = data_write,
.read = data_read,
.llseek = default_llseek,
};
static ssize_t size_read(struct file *file, char __user *ubuf, size_t count, loff_t *ppos)
{
struct dentry *dent = file_dentry(file)->d_parent;
struct xe_device *xe = extract_xe(dent);
unsigned int vfid = extract_vfid(dent);
char buf[21];
ssize_t ret;
int len;
xe_pm_runtime_get(xe);
ret = xe_sriov_pf_migration_size(xe, vfid);
xe_pm_runtime_put(xe);
if (ret < 0)
return ret;
len = scnprintf(buf, sizeof(buf), "%zd\n", ret);
return simple_read_from_buffer(ubuf, count, ppos, buf, len);
}
static const struct file_operations size_vf_fops = {
.owner = THIS_MODULE,
.open = simple_open,
.read = size_read,
.llseek = default_llseek,
};
static void pf_populate_vf(struct xe_device *xe, struct dentry *vfdent)
{
@ -190,6 +317,10 @@ static void pf_populate_vf(struct xe_device *xe, struct dentry *vfdent)
debugfs_create_file("resume", 0200, vfdent, xe, &resume_vf_fops);
debugfs_create_file("stop", 0200, vfdent, xe, &stop_vf_fops);
debugfs_create_file("reset", 0200, vfdent, xe, &reset_vf_fops);
debugfs_create_file("save", 0600, vfdent, xe, &save_vf_fops);
debugfs_create_file("restore", 0600, vfdent, xe, &restore_vf_fops);
debugfs_create_file("migration_data", 0600, vfdent, xe, &data_vf_fops);
debugfs_create_file("migration_size", 0400, vfdent, xe, &size_vf_fops);
}
static void pf_populate_with_tiles(struct xe_device *xe, struct dentry *dent, unsigned int vfid)

16
drivers/gpu/drm/xe/xe_sriov_pf_helpers.h
View File

@ -48,10 +48,26 @@ static inline unsigned int xe_sriov_pf_num_vfs(const struct xe_device *xe)
return pci_num_vf(to_pci_dev(xe->drm.dev));
}
/**
* xe_sriov_pf_admin_only() - Check if PF is mainly used for VFs administration.
* @xe: the PF &xe_device
*
* Return: True if PF is mainly used for VFs administration.
*/
static inline bool xe_sriov_pf_admin_only(const struct xe_device *xe)
{
return !xe->info.probe_display;
}
static inline struct mutex *xe_sriov_pf_master_mutex(struct xe_device *xe)
{
xe_assert(xe, IS_SRIOV_PF(xe));
return &xe->sriov.pf.master_lock;
}
int xe_sriov_pf_arm_guard(struct xe_device *xe, struct xe_guard *guard,
bool write, void *who);
void xe_sriov_pf_disarm_guard(struct xe_device *xe, struct xe_guard *guard,
bool write, void *who);
#endif

342
drivers/gpu/drm/xe/xe_sriov_pf_migration.c Normal file
View File

@ -0,0 +1,342 @@
// SPDX-License-Identifier: MIT
/*
* Copyright © 2025 Intel Corporation
*/
#include <drm/drm_managed.h>
#include "xe_device.h"
#include "xe_gt_sriov_pf_control.h"
#include "xe_gt_sriov_pf_migration.h"
#include "xe_pm.h"
#include "xe_sriov.h"
#include "xe_sriov_packet.h"
#include "xe_sriov_packet_types.h"
#include "xe_sriov_pf_helpers.h"
#include "xe_sriov_pf_migration.h"
#include "xe_sriov_printk.h"
static struct xe_sriov_migration_state *pf_pick_migration(struct xe_device *xe, unsigned int vfid)
{
xe_assert(xe, IS_SRIOV_PF(xe));
xe_assert(xe, vfid <= xe_sriov_pf_get_totalvfs(xe));
return &xe->sriov.pf.vfs[vfid].migration;
}
/**
* xe_sriov_pf_migration_waitqueue() - Get waitqueue for migration.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* Return: pointer to the migration waitqueue.
*/
wait_queue_head_t *xe_sriov_pf_migration_waitqueue(struct xe_device *xe, unsigned int vfid)
{
return &pf_pick_migration(xe, vfid)->wq;
}
/**
* xe_sriov_pf_migration_supported() - Check if SR-IOV VF migration is supported by the device
* @xe: the &xe_device
*
* Return: true if migration is supported, false otherwise
*/
bool xe_sriov_pf_migration_supported(struct xe_device *xe)
{
xe_assert(xe, IS_SRIOV_PF(xe));
return xe->sriov.pf.migration.supported;
}
static bool pf_check_migration_support(struct xe_device *xe)
{
/* XXX: for now this is for feature enabling only */
return IS_ENABLED(CONFIG_DRM_XE_DEBUG);
}
static void pf_migration_cleanup(void *arg)
{
struct xe_sriov_migration_state *migration = arg;
xe_sriov_packet_free(migration->pending);
xe_sriov_packet_free(migration->trailer);
xe_sriov_packet_free(migration->descriptor);
}
/**
* xe_sriov_pf_migration_init() - Initialize support for SR-IOV VF migration.
* @xe: the &xe_device
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_migration_init(struct xe_device *xe)
{
unsigned int n, totalvfs;
int err;
xe_assert(xe, IS_SRIOV_PF(xe));
xe->sriov.pf.migration.supported = pf_check_migration_support(xe);
if (!xe_sriov_pf_migration_supported(xe))
return 0;
totalvfs = xe_sriov_pf_get_totalvfs(xe);
for (n = 1; n <= totalvfs; n++) {
struct xe_sriov_migration_state *migration = pf_pick_migration(xe, n);
err = drmm_mutex_init(&xe->drm, &migration->lock);
if (err)
return err;
init_waitqueue_head(&migration->wq);
err = devm_add_action_or_reset(xe->drm.dev, pf_migration_cleanup, migration);
if (err)
return err;
}
return 0;
}
static bool pf_migration_data_ready(struct xe_device *xe, unsigned int vfid)
{
struct xe_gt *gt;
u8 gt_id;
for_each_gt(gt, xe, gt_id) {
if (xe_gt_sriov_pf_control_check_save_failed(gt, vfid) ||
xe_gt_sriov_pf_control_check_save_data_done(gt, vfid) ||
!xe_gt_sriov_pf_migration_ring_empty(gt, vfid))
return true;
}
return false;
}
static struct xe_sriov_packet *
pf_migration_consume(struct xe_device *xe, unsigned int vfid)
{
struct xe_sriov_packet *data;
bool more_data = false;
struct xe_gt *gt;
u8 gt_id;
for_each_gt(gt, xe, gt_id) {
data = xe_gt_sriov_pf_migration_save_consume(gt, vfid);
if (data && PTR_ERR(data) != EAGAIN)
return data;
if (PTR_ERR(data) == -EAGAIN)
more_data = true;
}
if (!more_data)
return NULL;
return ERR_PTR(-EAGAIN);
}
/**
* xe_sriov_pf_migration_save_consume() - Consume a VF migration data packet from the device.
* @xe: the &xe_device
* @vfid: the VF identifier
*
* Called by the save migration data consumer (userspace) when
* processing migration data.
* If there is no migration data to process, wait until more data is available.
*
* Return: Pointer to &xe_sriov_packet on success,
* NULL if ring is empty and no more migration data is expected,
* ERR_PTR value in case of error.
*
* Return: 0 on success or a negative error code on failure.
*/
struct xe_sriov_packet *
xe_sriov_pf_migration_save_consume(struct xe_device *xe, unsigned int vfid)
{
struct xe_sriov_migration_state *migration = pf_pick_migration(xe, vfid);
struct xe_sriov_packet *data;
int ret;
xe_assert(xe, IS_SRIOV_PF(xe));
for (;;) {
data = pf_migration_consume(xe, vfid);
if (PTR_ERR(data) != -EAGAIN)
break;
ret = wait_event_interruptible(migration->wq,
pf_migration_data_ready(xe, vfid));
if (ret)
return ERR_PTR(ret);
}
return data;
}
static int pf_handle_descriptor(struct xe_device *xe, unsigned int vfid,
struct xe_sriov_packet *data)
{
int ret;
if (data->hdr.tile_id != 0 || data->hdr.gt_id != 0)
return -EINVAL;
ret = xe_sriov_packet_process_descriptor(xe, vfid, data);
if (ret)
return ret;
xe_sriov_packet_free(data);
return 0;
}
static int pf_handle_trailer(struct xe_device *xe, unsigned int vfid,
struct xe_sriov_packet *data)
{
struct xe_gt *gt;
u8 gt_id;
if (data->hdr.tile_id != 0 || data->hdr.gt_id != 0)
return -EINVAL;
if (data->hdr.offset != 0 || data->hdr.size != 0 || data->buff || data->bo)
return -EINVAL;
xe_sriov_packet_free(data);
for_each_gt(gt, xe, gt_id)
xe_gt_sriov_pf_control_restore_data_done(gt, vfid);
return 0;
}
/**
* xe_sriov_pf_migration_restore_produce() - Produce a VF migration data packet to the device.
* @xe: the &xe_device
* @vfid: the VF identifier
* @data: Pointer to &xe_sriov_packet
*
* Called by the restore migration data producer (userspace) when processing
* migration data.
* If the underlying data structure is full, wait until there is space.
*
* Return: 0 on success or a negative error code on failure.
*/
int xe_sriov_pf_migration_restore_produce(struct xe_device *xe, unsigned int vfid,
struct xe_sriov_packet *data)
{
struct xe_gt *gt;
xe_assert(xe, IS_SRIOV_PF(xe));
if (data->hdr.type == XE_SRIOV_PACKET_TYPE_DESCRIPTOR)
return pf_handle_descriptor(xe, vfid, data);
if (data->hdr.type == XE_SRIOV_PACKET_TYPE_TRAILER)
return pf_handle_trailer(xe, vfid, data);
gt = xe_device_get_gt(xe, data->hdr.gt_id);
if (!gt || data->hdr.tile_id != gt->tile->id || data->hdr.type == 0) {
xe_sriov_err_ratelimited(xe, "Received invalid restore packet for VF%u (type:%u, tile:%u, GT:%u)\n",
vfid, data->hdr.type, data->hdr.tile_id, data->hdr.gt_id);
return -EINVAL;
}
return xe_gt_sriov_pf_migration_restore_produce(gt, vfid, data);
}
/**
* xe_sriov_pf_migration_read() - Read migration data from the device.
* @xe: the &xe_device
* @vfid: the VF identifier
* @buf: start address of userspace buffer
* @len: requested read size from userspace
*
* Return: number of bytes that has been successfully read,
* 0 if no more migration data is available,
* -errno on failure.
*/
ssize_t xe_sriov_pf_migration_read(struct xe_device *xe, unsigned int vfid,
char __user *buf, size_t len)
{
struct xe_sriov_migration_state *migration = pf_pick_migration(xe, vfid);
ssize_t ret, consumed = 0;
xe_assert(xe, IS_SRIOV_PF(xe));
scoped_cond_guard(mutex_intr, return -EINTR, &migration->lock) {
while (consumed < len) {
ret = xe_sriov_packet_read_single(xe, vfid, buf, len - consumed);
if (ret == -ENODATA)
break;
if (ret < 0)
return ret;
consumed += ret;
buf += ret;
}
}
return consumed;
}
/**
* xe_sriov_pf_migration_write() - Write migration data to the device.
* @xe: the &xe_device
* @vfid: the VF identifier
* @buf: start address of userspace buffer
* @len: requested write size from userspace
*
* Return: number of bytes that has been successfully written,
* -errno on failure.
*/
ssize_t xe_sriov_pf_migration_write(struct xe_device *xe, unsigned int vfid,
const char __user *buf, size_t len)
{
struct xe_sriov_migration_state *migration = pf_pick_migration(xe, vfid);
ssize_t ret, produced = 0;
xe_assert(xe, IS_SRIOV_PF(xe));
scoped_cond_guard(mutex_intr, return -EINTR, &migration->lock) {
while (produced < len) {
ret = xe_sriov_packet_write_single(xe, vfid, buf, len - produced);
if (ret < 0)
return ret;
produced += ret;
buf += ret;
}
}
return produced;
}
/**
* xe_sriov_pf_migration_size() - Total size of migration data from all components within a device
* @xe: the &xe_device
* @vfid: the VF identifier (can't be 0)
*
* This function is for PF only.
*
* Return: total migration data size in bytes or a negative error code on failure.
*/
ssize_t xe_sriov_pf_migration_size(struct xe_device *xe, unsigned int vfid)
{
size_t size = 0;
struct xe_gt *gt;
ssize_t ret;
u8 gt_id;
xe_assert(xe, IS_SRIOV_PF(xe));
xe_assert(xe, vfid);
for_each_gt(gt, xe, gt_id) {
ret = xe_gt_sriov_pf_migration_size(gt, vfid);
if (ret < 0)
return ret;
size += ret;
}
return size;
}

29
drivers/gpu/drm/xe/xe_sriov_pf_migration.h Normal file
View File

@ -0,0 +1,29 @@
/* SPDX-License-Identifier: MIT */
/*
* Copyright © 2025 Intel Corporation
*/
#ifndef _XE_SRIOV_PF_MIGRATION_H_
#define _XE_SRIOV_PF_MIGRATION_H_
#include <linux/types.h>
#include <linux/wait.h>
struct xe_device;
struct xe_sriov_packet;
int xe_sriov_pf_migration_init(struct xe_device *xe);
bool xe_sriov_pf_migration_supported(struct xe_device *xe);
int xe_sriov_pf_migration_restore_produce(struct xe_device *xe, unsigned int vfid,
struct xe_sriov_packet *data);
struct xe_sriov_packet *
xe_sriov_pf_migration_save_consume(struct xe_device *xe, unsigned int vfid);
ssize_t xe_sriov_pf_migration_size(struct xe_device *xe, unsigned int vfid);
wait_queue_head_t *xe_sriov_pf_migration_waitqueue(struct xe_device *xe, unsigned int vfid);
ssize_t xe_sriov_pf_migration_read(struct xe_device *xe, unsigned int vfid,
char __user *buf, size_t len);
ssize_t xe_sriov_pf_migration_write(struct xe_device *xe, unsigned int vfid,
const char __user *buf, size_t len);
#endif

37
drivers/gpu/drm/xe/xe_sriov_pf_migration_types.h Normal file
View File

@ -0,0 +1,37 @@
/* SPDX-License-Identifier: MIT */
/*
* Copyright © 2025 Intel Corporation
*/
#ifndef _XE_SRIOV_PF_MIGRATION_TYPES_H_
#define _XE_SRIOV_PF_MIGRATION_TYPES_H_
#include <linux/types.h>
#include <linux/mutex_types.h>
#include <linux/wait.h>
/**
* struct xe_sriov_pf_migration - Xe device level VF migration data
*/
struct xe_sriov_pf_migration {
/** @supported: indicates whether VF migration feature is supported */
bool supported;
};
/**
* struct xe_sriov_migration_state - Per VF device-level migration related data
*/
struct xe_sriov_migration_state {
/** @wq: waitqueue used to avoid busy-waiting for snapshot production/consumption */
wait_queue_head_t wq;
/** @lock: Mutex protecting the migration data */
struct mutex lock;
/** @pending: currently processed data packet of VF resource */
struct xe_sriov_packet *pending;
/** @trailer: data packet used to indicate the end of stream */
struct xe_sriov_packet *trailer;
/** @descriptor: data packet containing the metadata describing the device */
struct xe_sriov_packet *descriptor;
};
#endif

10
drivers/gpu/drm/xe/xe_sriov_pf_types.h
View File

@ -9,6 +9,8 @@
#include <linux/mutex.h>
#include <linux/types.h>
#include "xe_guard.h"
#include "xe_sriov_pf_migration_types.h"
#include "xe_sriov_pf_provision_types.h"
#include "xe_sriov_pf_service_types.h"
@ -23,6 +25,8 @@ struct xe_sriov_metadata {
/** @version: negotiated VF/PF ABI version */
struct xe_sriov_pf_service_version version;
/** @migration: migration state */
struct xe_sriov_migration_state migration;
};
/**
@ -38,12 +42,18 @@ struct xe_device_pf {
/** @driver_max_vfs: Maximum number of VFs supported by the driver. */
u16 driver_max_vfs;
/** @guard_vfs_enabling: guards VFs enabling */
struct xe_guard guard_vfs_enabling;
/** @master_lock: protects all VFs configurations across GTs */
struct mutex master_lock;
/** @provision: device level provisioning data. */
struct xe_sriov_pf_provision provision;
/** @migration: device level migration data. */
struct xe_sriov_pf_migration migration;
/** @service: device level service data. */
struct xe_sriov_pf_service service;

8
drivers/gpu/drm/xe/xe_sriov_vf.c
View File

@ -156,14 +156,6 @@ void xe_sriov_vf_migration_disable(struct xe_device *xe, const char *fmt, ...)
static void vf_migration_init_early(struct xe_device *xe)
{
/*
* TODO: Add conditions to allow specific platforms, when they're
* supported at production quality.
*/
if (!IS_ENABLED(CONFIG_DRM_XE_DEBUG))
return xe_sriov_vf_migration_disable(xe,
"experimental feature not available on production builds");
if (!xe_device_has_memirq(xe))
return xe_sriov_vf_migration_disable(xe, "requires memory-based IRQ support");

26
drivers/gpu/drm/xe/xe_vram.c
View File

@ -183,12 +183,17 @@ static int determine_lmem_bar_size(struct xe_device *xe, struct xe_vram_region *
return 0;
}
static inline u64 get_flat_ccs_offset(struct xe_gt *gt, u64 tile_size)
static int get_flat_ccs_offset(struct xe_gt *gt, u64 tile_size, u64 *poffset)
{
struct xe_device *xe = gt_to_xe(gt);
unsigned int fw_ref;
u64 offset;
u32 reg;
fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FW_GT);
if (!fw_ref)
return -ETIMEDOUT;
if (GRAPHICS_VER(xe) >= 20) {
u64 ccs_size = tile_size / 512;
u64 offset_hi, offset_lo;
@ -218,7 +223,10 @@ static inline u64 get_flat_ccs_offset(struct xe_gt *gt, u64 tile_size)
offset = (u64)REG_FIELD_GET(XEHP_FLAT_CCS_PTR, reg) * SZ_64K;
}
return offset;
xe_force_wake_put(gt_to_fw(gt), fw_ref);
*poffset = offset;
return 0;
}
/*
@ -245,7 +253,6 @@ static int tile_vram_size(struct xe_tile *tile, u64 *vram_size,
{
struct xe_device *xe = tile_to_xe(tile);
struct xe_gt *gt = tile->primary_gt;
unsigned int fw_ref;
u64 offset;
u32 reg;
@ -265,23 +272,22 @@ static int tile_vram_size(struct xe_tile *tile, u64 *vram_size,
return 0;
}
fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FW_GT);
if (!fw_ref)
return -ETIMEDOUT;
/* actual size */
if (unlikely(xe->info.platform == XE_DG1)) {
*tile_size = pci_resource_len(to_pci_dev(xe->drm.dev), LMEM_BAR);
*tile_offset = 0;
} else {
reg = xe_gt_mcr_unicast_read_any(gt, XEHP_TILE_ADDR_RANGE(gt->info.id));
reg = xe_mmio_read32(&tile->mmio, SG_TILE_ADDR_RANGE(tile->id));
*tile_size = (u64)REG_FIELD_GET(GENMASK(14, 8), reg) * SZ_1G;
*tile_offset = (u64)REG_FIELD_GET(GENMASK(7, 1), reg) * SZ_1G;
}
/* minus device usage */
if (xe->info.has_flat_ccs) {
offset = get_flat_ccs_offset(gt, *tile_size);
int ret = get_flat_ccs_offset(gt, *tile_size, &offset);
if (ret)
return ret;
} else {
offset = xe_mmio_read64_2x32(&tile->mmio, GSMBASE);
}
@ -289,8 +295,6 @@ static int tile_vram_size(struct xe_tile *tile, u64 *vram_size,
/* remove the tile offset so we have just the available size */
*vram_size = offset - *tile_offset;
xe_force_wake_put(gt_to_fw(gt), fw_ref);
return 0;
}

5
drivers/gpu/drm/xe/xe_wa.c
View File

@ -922,6 +922,11 @@ static const struct xe_rtp_entry_sr lrc_was[] = {
XE_RTP_RULES(GRAPHICS_VERSION_RANGE(3000, 3005), ENGINE_CLASS(RENDER)),
XE_RTP_ACTIONS(SET(XEHP_SLICE_COMMON_ECO_CHICKEN1, FAST_CLEAR_VALIGN_FIX))
},
{ XE_RTP_NAME("15016589081"),
XE_RTP_RULES(GRAPHICS_VERSION(3000), GRAPHICS_STEP(A0, B0),
ENGINE_CLASS(RENDER)),
XE_RTP_ACTIONS(SET(CHICKEN_RASTER_1, DIS_CLIP_NEGATIVE_BOUNDING_BOX))
},
};
static __maybe_unused const struct xe_rtp_entry oob_was[] = {

2
include/drm/intel/pciids.h
View File

@ -861,7 +861,7 @@
MACRO__(0xE216, ## __VA_ARGS__)
#define INTEL_BMG_IDS(MACRO__, ...) \
INTEL_BMG_G21_IDS(MACRO__, __VA_ARGS__), \
INTEL_BMG_G21_IDS(MACRO__, ## __VA_ARGS__), \
MACRO__(0xE220, ## __VA_ARGS__), \
MACRO__(0xE221, ## __VA_ARGS__), \
MACRO__(0xE222, ## __VA_ARGS__), \