Files
kernel_xiaomi_sm8250/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
Greg Kroah-Hartman d1253c75a8 Merge 4.19.155 into android-4.19-stable
Changes in 4.19.155
	objtool: Support Clang non-section symbols in ORC generation
	scripts/setlocalversion: make git describe output more reliable
	arm64: Run ARCH_WORKAROUND_1 enabling code on all CPUs
	arm64: link with -z norelro regardless of CONFIG_RELOCATABLE
	x86/PCI: Fix intel_mid_pci.c build error when ACPI is not enabled
	efivarfs: Replace invalid slashes with exclamation marks in dentries.
	chelsio/chtls: fix deadlock issue
	chelsio/chtls: fix memory leaks in CPL handlers
	chelsio/chtls: fix tls record info to user
	gtp: fix an use-before-init in gtp_newlink()
	mlxsw: core: Fix memory leak on module removal
	netem: fix zero division in tabledist
	ravb: Fix bit fields checking in ravb_hwtstamp_get()
	tcp: Prevent low rmem stalls with SO_RCVLOWAT.
	tipc: fix memory leak caused by tipc_buf_append()
	r8169: fix issue with forced threading in combination with shared interrupts
	cxgb4: set up filter action after rewrites
	arch/x86/amd/ibs: Fix re-arming IBS Fetch
	x86/xen: disable Firmware First mode for correctable memory errors
	fuse: fix page dereference after free
	bpf: Fix comment for helper bpf_current_task_under_cgroup()
	evm: Check size of security.evm before using it
	p54: avoid accessing the data mapped to streaming DMA
	cxl: Rework error message for incompatible slots
	RDMA/addr: Fix race with netevent_callback()/rdma_addr_cancel()
	mtd: lpddr: Fix bad logic in print_drs_error
	serial: pl011: Fix lockdep splat when handling magic-sysrq interrupt
	ata: sata_rcar: Fix DMA boundary mask
	fscrypt: return -EXDEV for incompatible rename or link into encrypted dir
	fscrypt: clean up and improve dentry revalidation
	fscrypt: fix race allowing rename() and link() of ciphertext dentries
	fs, fscrypt: clear DCACHE_ENCRYPTED_NAME when unaliasing directory
	fscrypt: only set dentry_operations on ciphertext dentries
	fscrypt: fix race where ->lookup() marks plaintext dentry as ciphertext
	Revert "block: ratelimit handle_bad_sector() message"
	xen/events: don't use chip_data for legacy IRQs
	xen/events: avoid removing an event channel while handling it
	xen/events: add a proper barrier to 2-level uevent unmasking
	xen/events: fix race in evtchn_fifo_unmask()
	xen/events: add a new "late EOI" evtchn framework
	xen/blkback: use lateeoi irq binding
	xen/netback: use lateeoi irq binding
	xen/scsiback: use lateeoi irq binding
	xen/pvcallsback: use lateeoi irq binding
	xen/pciback: use lateeoi irq binding
	xen/events: switch user event channels to lateeoi model
	xen/events: use a common cpu hotplug hook for event channels
	xen/events: defer eoi in case of excessive number of events
	xen/events: block rogue events for some time
	x86/unwind/orc: Fix inactive tasks with stack pointer in %sp on GCC 10 compiled kernels
	mlxsw: core: Fix use-after-free in mlxsw_emad_trans_finish()
	RDMA/qedr: Fix memory leak in iWARP CM
	ata: sata_nv: Fix retrieving of active qcs
	futex: Fix incorrect should_fail_futex() handling
	powerpc/powernv/smp: Fix spurious DBG() warning
	mm: fix exec activate_mm vs TLB shootdown and lazy tlb switching race
	powerpc: select ARCH_WANT_IRQS_OFF_ACTIVATE_MM
	sparc64: remove mm_cpumask clearing to fix kthread_use_mm race
	f2fs: add trace exit in exception path
	f2fs: fix uninit-value in f2fs_lookup
	f2fs: fix to check segment boundary during SIT page readahead
	um: change sigio_spinlock to a mutex
	ARM: 8997/2: hw_breakpoint: Handle inexact watchpoint addresses
	power: supply: bq27xxx: report "not charging" on all types
	xfs: fix realtime bitmap/summary file truncation when growing rt volume
	video: fbdev: pvr2fb: initialize variables
	ath10k: start recovery process when payload length exceeds max htc length for sdio
	ath10k: fix VHT NSS calculation when STBC is enabled
	drm/brige/megachips: Add checking if ge_b850v3_lvds_init() is working correctly
	media: videodev2.h: RGB BT2020 and HSV are always full range
	media: platform: Improve queue set up flow for bug fixing
	usb: typec: tcpm: During PR_SWAP, source caps should be sent only after tSwapSourceStart
	media: tw5864: check status of tw5864_frameinterval_get
	media: imx274: fix frame interval handling
	mmc: via-sdmmc: Fix data race bug
	drm/bridge/synopsys: dsi: add support for non-continuous HS clock
	arm64: topology: Stop using MPIDR for topology information
	printk: reduce LOG_BUF_SHIFT range for H8300
	ia64: kprobes: Use generic kretprobe trampoline handler
	kgdb: Make "kgdbcon" work properly with "kgdb_earlycon"
	media: uvcvideo: Fix dereference of out-of-bound list iterator
	riscv: Define AT_VECTOR_SIZE_ARCH for ARCH_DLINFO
	cpufreq: sti-cpufreq: add stih418 support
	USB: adutux: fix debugging
	uio: free uio id after uio file node is freed
	usb: xhci: omit duplicate actions when suspending a runtime suspended host.
	arm64/mm: return cpu_all_mask when node is NUMA_NO_NODE
	xfs: don't free rt blocks when we're doing a REMAP bunmapi call
	ACPI: Add out of bounds and numa_off protections to pxm_to_node()
	drivers/net/wan/hdlc_fr: Correctly handle special skb->protocol values
	bus/fsl_mc: Do not rely on caller to provide non NULL mc_io
	power: supply: test_power: add missing newlines when printing parameters by sysfs
	drm/amd/display: HDMI remote sink need mode validation for Linux
	btrfs: fix replace of seed device
	md/bitmap: md_bitmap_get_counter returns wrong blocks
	bnxt_en: Log unknown link speed appropriately.
	rpmsg: glink: Use complete_all for open states
	clk: ti: clockdomain: fix static checker warning
	net: 9p: initialize sun_server.sun_path to have addr's value only when addr is valid
	drivers: watchdog: rdc321x_wdt: Fix race condition bugs
	ext4: Detect already used quota file early
	gfs2: add validation checks for size of superblock
	cifs: handle -EINTR in cifs_setattr
	arm64: dts: renesas: ulcb: add full-pwr-cycle-in-suspend into eMMC nodes
	ARM: dts: omap4: Fix sgx clock rate for 4430
	memory: emif: Remove bogus debugfs error handling
	ARM: dts: s5pv210: remove DMA controller bus node name to fix dtschema warnings
	ARM: dts: s5pv210: move PMU node out of clock controller
	ARM: dts: s5pv210: remove dedicated 'audio-subsystem' node
	nbd: make the config put is called before the notifying the waiter
	sgl_alloc_order: fix memory leak
	nvme-rdma: fix crash when connect rejected
	md/raid5: fix oops during stripe resizing
	mmc: sdhci-acpi: AMDI0040: Set SDHCI_QUIRK2_PRESET_VALUE_BROKEN
	perf/x86/amd/ibs: Don't include randomized bits in get_ibs_op_count()
	perf/x86/amd/ibs: Fix raw sample data accumulation
	leds: bcm6328, bcm6358: use devres LED registering function
	media: uvcvideo: Fix uvc_ctrl_fixup_xu_info() not having any effect
	fs: Don't invalidate page buffers in block_write_full_page()
	NFS: fix nfs_path in case of a rename retry
	ACPI: button: fix handling lid state changes when input device closed
	ACPI / extlog: Check for RDMSR failure
	ACPI: video: use ACPI backlight for HP 635 Notebook
	ACPI: debug: don't allow debugging when ACPI is disabled
	acpi-cpufreq: Honor _PSD table setting on new AMD CPUs
	w1: mxc_w1: Fix timeout resolution problem leading to bus error
	scsi: mptfusion: Fix null pointer dereferences in mptscsih_remove()
	scsi: qla2xxx: Fix crash on session cleanup with unload
	btrfs: qgroup: fix wrong qgroup metadata reserve for delayed inode
	btrfs: improve device scanning messages
	btrfs: reschedule if necessary when logging directory items
	btrfs: send, recompute reference path after orphanization of a directory
	btrfs: use kvzalloc() to allocate clone_roots in btrfs_ioctl_send()
	btrfs: cleanup cow block on error
	btrfs: fix use-after-free on readahead extent after failure to create it
	usb: xhci: Workaround for S3 issue on AMD SNPS 3.0 xHC
	usb: dwc3: ep0: Fix ZLP for OUT ep0 requests
	usb: dwc3: gadget: Check MPS of the request length
	usb: dwc3: core: add phy cleanup for probe error handling
	usb: dwc3: core: don't trigger runtime pm when remove driver
	usb: cdc-acm: fix cooldown mechanism
	usb: typec: tcpm: reset hard_reset_count for any disconnect
	usb: host: fsl-mph-dr-of: check return of dma_set_mask()
	drm/i915: Force VT'd workarounds when running as a guest OS
	vt: keyboard, simplify vt_kdgkbsent
	vt: keyboard, extend func_buf_lock to readers
	HID: wacom: Avoid entering wacom_wac_pen_report for pad / battery
	udf: Fix memory leak when mounting
	dmaengine: dma-jz4780: Fix race in jz4780_dma_tx_status
	iio:light:si1145: Fix timestamp alignment and prevent data leak.
	iio:adc:ti-adc0832 Fix alignment issue with timestamp
	iio:adc:ti-adc12138 Fix alignment issue with timestamp
	iio:gyro:itg3200: Fix timestamp alignment and prevent data leak.
	powerpc/drmem: Make lmb_size 64 bit
	s390/stp: add locking to sysfs functions
	powerpc/rtas: Restrict RTAS requests from userspace
	powerpc: Warn about use of smt_snooze_delay
	powerpc/powernv/elog: Fix race while processing OPAL error log event.
	powerpc: Fix undetected data corruption with P9N DD2.1 VSX CI load emulation
	NFSv4.2: support EXCHGID4_FLAG_SUPP_FENCE_OPS 4.2 EXCHANGE_ID flag
	NFSD: Add missing NFSv2 .pc_func methods
	ubifs: dent: Fix some potential memory leaks while iterating entries
	perf python scripting: Fix printable strings in python3 scripts
	ubi: check kthread_should_stop() after the setting of task state
	ia64: fix build error with !COREDUMP
	i2c: imx: Fix external abort on interrupt in exit paths
	drm/amdgpu: don't map BO in reserved region
	drm/amd/display: Don't invoke kgdb_breakpoint() unconditionally
	ceph: promote to unsigned long long before shifting
	libceph: clear con->out_msg on Policy::stateful_server faults
	9P: Cast to loff_t before multiplying
	ring-buffer: Return 0 on success from ring_buffer_resize()
	vringh: fix __vringh_iov() when riov and wiov are different
	ext4: fix leaking sysfs kobject after failed mount
	ext4: fix error handling code in add_new_gdb
	ext4: fix invalid inode checksum
	drm/ttm: fix eviction valuable range check.
	rtc: rx8010: don't modify the global rtc ops
	tty: make FONTX ioctl use the tty pointer they were actually passed
	arm64: berlin: Select DW_APB_TIMER_OF
	cachefiles: Handle readpage error correctly
	hil/parisc: Disable HIL driver when it gets stuck
	arm: dts: mt7623: add missing pause for switchport
	ARM: samsung: fix PM debug build with DEBUG_LL but !MMU
	ARM: s3c24xx: fix missing system reset
	device property: Keep secondary firmware node secondary by type
	device property: Don't clear secondary pointer for shared primary firmware node
	KVM: arm64: Fix AArch32 handling of DBGD{CCINT,SCRext} and DBGVCR
	staging: comedi: cb_pcidas: Allow 2-channel commands for AO subdevice
	staging: octeon: repair "fixed-link" support
	staging: octeon: Drop on uncorrectable alignment or FCS error
	Linux 4.19.155

Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I18fefb5bfaa4d05772c61c2975340d0f089b8e3e
2020-11-05 14:02:27 +01:00

903 lines
23 KiB
C

/*
* Copyright 2008 Advanced Micro Devices, Inc.
* Copyright 2008 Red Hat Inc.
* Copyright 2009 Jerome Glisse.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
* Authors: Dave Airlie
* Alex Deucher
* Jerome Glisse
*/
#include <linux/ktime.h>
#include <linux/pagemap.h>
#include <drm/drmP.h>
#include <drm/amdgpu_drm.h>
#include "amdgpu.h"
#include "amdgpu_display.h"
void amdgpu_gem_object_free(struct drm_gem_object *gobj)
{
struct amdgpu_bo *robj = gem_to_amdgpu_bo(gobj);
if (robj) {
amdgpu_mn_unregister(robj);
amdgpu_bo_unref(&robj);
}
}
int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
int alignment, u32 initial_domain,
u64 flags, enum ttm_bo_type type,
struct reservation_object *resv,
struct drm_gem_object **obj)
{
struct amdgpu_bo *bo;
struct amdgpu_bo_param bp;
int r;
memset(&bp, 0, sizeof(bp));
*obj = NULL;
/* At least align on page size */
if (alignment < PAGE_SIZE) {
alignment = PAGE_SIZE;
}
bp.size = size;
bp.byte_align = alignment;
bp.type = type;
bp.resv = resv;
bp.preferred_domain = initial_domain;
retry:
bp.flags = flags;
bp.domain = initial_domain;
r = amdgpu_bo_create(adev, &bp, &bo);
if (r) {
if (r != -ERESTARTSYS) {
if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
goto retry;
}
if (initial_domain == AMDGPU_GEM_DOMAIN_VRAM) {
initial_domain |= AMDGPU_GEM_DOMAIN_GTT;
goto retry;
}
DRM_DEBUG("Failed to allocate GEM object (%ld, %d, %u, %d)\n",
size, initial_domain, alignment, r);
}
return r;
}
*obj = &bo->gem_base;
return 0;
}
void amdgpu_gem_force_release(struct amdgpu_device *adev)
{
struct drm_device *ddev = adev->ddev;
struct drm_file *file;
mutex_lock(&ddev->filelist_mutex);
list_for_each_entry(file, &ddev->filelist, lhead) {
struct drm_gem_object *gobj;
int handle;
WARN_ONCE(1, "Still active user space clients!\n");
spin_lock(&file->table_lock);
idr_for_each_entry(&file->object_idr, gobj, handle) {
WARN_ONCE(1, "And also active allocations!\n");
drm_gem_object_put_unlocked(gobj);
}
idr_destroy(&file->object_idr);
spin_unlock(&file->table_lock);
}
mutex_unlock(&ddev->filelist_mutex);
}
/*
* Call from drm_gem_handle_create which appear in both new and open ioctl
* case.
*/
int amdgpu_gem_object_open(struct drm_gem_object *obj,
struct drm_file *file_priv)
{
struct amdgpu_bo *abo = gem_to_amdgpu_bo(obj);
struct amdgpu_device *adev = amdgpu_ttm_adev(abo->tbo.bdev);
struct amdgpu_fpriv *fpriv = file_priv->driver_priv;
struct amdgpu_vm *vm = &fpriv->vm;
struct amdgpu_bo_va *bo_va;
struct mm_struct *mm;
int r;
mm = amdgpu_ttm_tt_get_usermm(abo->tbo.ttm);
if (mm && mm != current->mm)
return -EPERM;
if (abo->flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID &&
abo->tbo.resv != vm->root.base.bo->tbo.resv)
return -EPERM;
r = amdgpu_bo_reserve(abo, false);
if (r)
return r;
bo_va = amdgpu_vm_bo_find(vm, abo);
if (!bo_va) {
bo_va = amdgpu_vm_bo_add(adev, vm, abo);
} else {
++bo_va->ref_count;
}
amdgpu_bo_unreserve(abo);
return 0;
}
void amdgpu_gem_object_close(struct drm_gem_object *obj,
struct drm_file *file_priv)
{
struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
struct amdgpu_fpriv *fpriv = file_priv->driver_priv;
struct amdgpu_vm *vm = &fpriv->vm;
struct amdgpu_bo_list_entry vm_pd;
struct list_head list, duplicates;
struct ttm_validate_buffer tv;
struct ww_acquire_ctx ticket;
struct amdgpu_bo_va *bo_va;
int r;
INIT_LIST_HEAD(&list);
INIT_LIST_HEAD(&duplicates);
tv.bo = &bo->tbo;
tv.shared = true;
list_add(&tv.head, &list);
amdgpu_vm_get_pd_bo(vm, &list, &vm_pd);
r = ttm_eu_reserve_buffers(&ticket, &list, false, &duplicates);
if (r) {
dev_err(adev->dev, "leaking bo va because "
"we fail to reserve bo (%d)\n", r);
return;
}
bo_va = amdgpu_vm_bo_find(vm, bo);
if (bo_va && --bo_va->ref_count == 0) {
amdgpu_vm_bo_rmv(adev, bo_va);
if (amdgpu_vm_ready(vm)) {
struct dma_fence *fence = NULL;
r = amdgpu_vm_clear_freed(adev, vm, &fence);
if (unlikely(r)) {
dev_err(adev->dev, "failed to clear page "
"tables on GEM object close (%d)\n", r);
}
if (fence) {
amdgpu_bo_fence(bo, fence, true);
dma_fence_put(fence);
}
}
}
ttm_eu_backoff_reservation(&ticket, &list);
}
/*
* GEM ioctls.
*/
int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
struct amdgpu_device *adev = dev->dev_private;
struct amdgpu_fpriv *fpriv = filp->driver_priv;
struct amdgpu_vm *vm = &fpriv->vm;
union drm_amdgpu_gem_create *args = data;
uint64_t flags = args->in.domain_flags;
uint64_t size = args->in.bo_size;
struct reservation_object *resv = NULL;
struct drm_gem_object *gobj;
uint32_t handle;
int r;
/* reject invalid gem flags */
if (flags & ~(AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
AMDGPU_GEM_CREATE_NO_CPU_ACCESS |
AMDGPU_GEM_CREATE_CPU_GTT_USWC |
AMDGPU_GEM_CREATE_VRAM_CLEARED |
AMDGPU_GEM_CREATE_VM_ALWAYS_VALID |
AMDGPU_GEM_CREATE_EXPLICIT_SYNC))
return -EINVAL;
/* reject invalid gem domains */
if (args->in.domains & ~AMDGPU_GEM_DOMAIN_MASK)
return -EINVAL;
/* create a gem object to contain this object in */
if (args->in.domains & (AMDGPU_GEM_DOMAIN_GDS |
AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA)) {
if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {
/* if gds bo is created from user space, it must be
* passed to bo list
*/
DRM_ERROR("GDS bo cannot be per-vm-bo\n");
return -EINVAL;
}
flags |= AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
if (args->in.domains == AMDGPU_GEM_DOMAIN_GDS)
size = size << AMDGPU_GDS_SHIFT;
else if (args->in.domains == AMDGPU_GEM_DOMAIN_GWS)
size = size << AMDGPU_GWS_SHIFT;
else if (args->in.domains == AMDGPU_GEM_DOMAIN_OA)
size = size << AMDGPU_OA_SHIFT;
else
return -EINVAL;
}
size = roundup(size, PAGE_SIZE);
if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {
r = amdgpu_bo_reserve(vm->root.base.bo, false);
if (r)
return r;
resv = vm->root.base.bo->tbo.resv;
}
r = amdgpu_gem_object_create(adev, size, args->in.alignment,
(u32)(0xffffffff & args->in.domains),
flags, ttm_bo_type_device, resv, &gobj);
if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {
if (!r) {
struct amdgpu_bo *abo = gem_to_amdgpu_bo(gobj);
abo->parent = amdgpu_bo_ref(vm->root.base.bo);
}
amdgpu_bo_unreserve(vm->root.base.bo);
}
if (r)
return r;
r = drm_gem_handle_create(filp, gobj, &handle);
/* drop reference from allocate - handle holds it now */
drm_gem_object_put_unlocked(gobj);
if (r)
return r;
memset(args, 0, sizeof(*args));
args->out.handle = handle;
return 0;
}
int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
struct ttm_operation_ctx ctx = { true, false };
struct amdgpu_device *adev = dev->dev_private;
struct drm_amdgpu_gem_userptr *args = data;
struct drm_gem_object *gobj;
struct amdgpu_bo *bo;
uint32_t handle;
int r;
args->addr = untagged_addr(args->addr);
if (offset_in_page(args->addr | args->size))
return -EINVAL;
/* reject unknown flag values */
if (args->flags & ~(AMDGPU_GEM_USERPTR_READONLY |
AMDGPU_GEM_USERPTR_ANONONLY | AMDGPU_GEM_USERPTR_VALIDATE |
AMDGPU_GEM_USERPTR_REGISTER))
return -EINVAL;
if (!(args->flags & AMDGPU_GEM_USERPTR_READONLY) &&
!(args->flags & AMDGPU_GEM_USERPTR_REGISTER)) {
/* if we want to write to it we must install a MMU notifier */
return -EACCES;
}
/* create a gem object to contain this object in */
r = amdgpu_gem_object_create(adev, args->size, 0, AMDGPU_GEM_DOMAIN_CPU,
0, ttm_bo_type_device, NULL, &gobj);
if (r)
return r;
bo = gem_to_amdgpu_bo(gobj);
bo->preferred_domains = AMDGPU_GEM_DOMAIN_GTT;
bo->allowed_domains = AMDGPU_GEM_DOMAIN_GTT;
r = amdgpu_ttm_tt_set_userptr(bo->tbo.ttm, args->addr, args->flags);
if (r)
goto release_object;
if (args->flags & AMDGPU_GEM_USERPTR_REGISTER) {
r = amdgpu_mn_register(bo, args->addr);
if (r)
goto release_object;
}
if (args->flags & AMDGPU_GEM_USERPTR_VALIDATE) {
r = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm,
bo->tbo.ttm->pages);
if (r)
goto release_object;
r = amdgpu_bo_reserve(bo, true);
if (r)
goto free_pages;
amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_GTT);
r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
amdgpu_bo_unreserve(bo);
if (r)
goto free_pages;
}
r = drm_gem_handle_create(filp, gobj, &handle);
/* drop reference from allocate - handle holds it now */
drm_gem_object_put_unlocked(gobj);
if (r)
return r;
args->handle = handle;
return 0;
free_pages:
release_pages(bo->tbo.ttm->pages, bo->tbo.ttm->num_pages);
release_object:
drm_gem_object_put_unlocked(gobj);
return r;
}
int amdgpu_mode_dumb_mmap(struct drm_file *filp,
struct drm_device *dev,
uint32_t handle, uint64_t *offset_p)
{
struct drm_gem_object *gobj;
struct amdgpu_bo *robj;
gobj = drm_gem_object_lookup(filp, handle);
if (gobj == NULL) {
return -ENOENT;
}
robj = gem_to_amdgpu_bo(gobj);
if (amdgpu_ttm_tt_get_usermm(robj->tbo.ttm) ||
(robj->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)) {
drm_gem_object_put_unlocked(gobj);
return -EPERM;
}
*offset_p = amdgpu_bo_mmap_offset(robj);
drm_gem_object_put_unlocked(gobj);
return 0;
}
int amdgpu_gem_mmap_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
union drm_amdgpu_gem_mmap *args = data;
uint32_t handle = args->in.handle;
memset(args, 0, sizeof(*args));
return amdgpu_mode_dumb_mmap(filp, dev, handle, &args->out.addr_ptr);
}
/**
* amdgpu_gem_timeout - calculate jiffies timeout from absolute value
*
* @timeout_ns: timeout in ns
*
* Calculate the timeout in jiffies from an absolute timeout in ns.
*/
unsigned long amdgpu_gem_timeout(uint64_t timeout_ns)
{
unsigned long timeout_jiffies;
ktime_t timeout;
/* clamp timeout if it's to large */
if (((int64_t)timeout_ns) < 0)
return MAX_SCHEDULE_TIMEOUT;
timeout = ktime_sub(ns_to_ktime(timeout_ns), ktime_get());
if (ktime_to_ns(timeout) < 0)
return 0;
timeout_jiffies = nsecs_to_jiffies(ktime_to_ns(timeout));
/* clamp timeout to avoid unsigned-> signed overflow */
if (timeout_jiffies > MAX_SCHEDULE_TIMEOUT )
return MAX_SCHEDULE_TIMEOUT - 1;
return timeout_jiffies;
}
int amdgpu_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
union drm_amdgpu_gem_wait_idle *args = data;
struct drm_gem_object *gobj;
struct amdgpu_bo *robj;
uint32_t handle = args->in.handle;
unsigned long timeout = amdgpu_gem_timeout(args->in.timeout);
int r = 0;
long ret;
gobj = drm_gem_object_lookup(filp, handle);
if (gobj == NULL) {
return -ENOENT;
}
robj = gem_to_amdgpu_bo(gobj);
ret = reservation_object_wait_timeout_rcu(robj->tbo.resv, true, true,
timeout);
/* ret == 0 means not signaled,
* ret > 0 means signaled
* ret < 0 means interrupted before timeout
*/
if (ret >= 0) {
memset(args, 0, sizeof(*args));
args->out.status = (ret == 0);
} else
r = ret;
drm_gem_object_put_unlocked(gobj);
return r;
}
int amdgpu_gem_metadata_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
struct drm_amdgpu_gem_metadata *args = data;
struct drm_gem_object *gobj;
struct amdgpu_bo *robj;
int r = -1;
DRM_DEBUG("%d \n", args->handle);
gobj = drm_gem_object_lookup(filp, args->handle);
if (gobj == NULL)
return -ENOENT;
robj = gem_to_amdgpu_bo(gobj);
r = amdgpu_bo_reserve(robj, false);
if (unlikely(r != 0))
goto out;
if (args->op == AMDGPU_GEM_METADATA_OP_GET_METADATA) {
amdgpu_bo_get_tiling_flags(robj, &args->data.tiling_info);
r = amdgpu_bo_get_metadata(robj, args->data.data,
sizeof(args->data.data),
&args->data.data_size_bytes,
&args->data.flags);
} else if (args->op == AMDGPU_GEM_METADATA_OP_SET_METADATA) {
if (args->data.data_size_bytes > sizeof(args->data.data)) {
r = -EINVAL;
goto unreserve;
}
r = amdgpu_bo_set_tiling_flags(robj, args->data.tiling_info);
if (!r)
r = amdgpu_bo_set_metadata(robj, args->data.data,
args->data.data_size_bytes,
args->data.flags);
}
unreserve:
amdgpu_bo_unreserve(robj);
out:
drm_gem_object_put_unlocked(gobj);
return r;
}
/**
* amdgpu_gem_va_update_vm -update the bo_va in its VM
*
* @adev: amdgpu_device pointer
* @vm: vm to update
* @bo_va: bo_va to update
* @operation: map, unmap or clear
*
* Update the bo_va directly after setting its address. Errors are not
* vital here, so they are not reported back to userspace.
*/
static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct amdgpu_bo_va *bo_va,
uint32_t operation)
{
int r;
if (!amdgpu_vm_ready(vm))
return;
r = amdgpu_vm_clear_freed(adev, vm, NULL);
if (r)
goto error;
if (operation == AMDGPU_VA_OP_MAP ||
operation == AMDGPU_VA_OP_REPLACE) {
r = amdgpu_vm_bo_update(adev, bo_va, false);
if (r)
goto error;
}
r = amdgpu_vm_update_directories(adev, vm);
error:
if (r && r != -ERESTARTSYS)
DRM_ERROR("Couldn't update BO_VA (%d)\n", r);
}
int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
const uint32_t valid_flags = AMDGPU_VM_DELAY_UPDATE |
AMDGPU_VM_PAGE_READABLE | AMDGPU_VM_PAGE_WRITEABLE |
AMDGPU_VM_PAGE_EXECUTABLE | AMDGPU_VM_MTYPE_MASK;
const uint32_t prt_flags = AMDGPU_VM_DELAY_UPDATE |
AMDGPU_VM_PAGE_PRT;
struct drm_amdgpu_gem_va *args = data;
struct drm_gem_object *gobj;
struct amdgpu_device *adev = dev->dev_private;
struct amdgpu_fpriv *fpriv = filp->driver_priv;
struct amdgpu_bo *abo;
struct amdgpu_bo_va *bo_va;
struct amdgpu_bo_list_entry vm_pd;
struct ttm_validate_buffer tv;
struct ww_acquire_ctx ticket;
struct list_head list, duplicates;
uint64_t va_flags;
uint64_t vm_size;
int r = 0;
if (args->va_address < AMDGPU_VA_RESERVED_SIZE) {
dev_dbg(&dev->pdev->dev,
"va_address 0x%LX is in reserved area 0x%LX\n",
args->va_address, AMDGPU_VA_RESERVED_SIZE);
return -EINVAL;
}
if (args->va_address >= AMDGPU_VA_HOLE_START &&
args->va_address < AMDGPU_VA_HOLE_END) {
dev_dbg(&dev->pdev->dev,
"va_address 0x%LX is in VA hole 0x%LX-0x%LX\n",
args->va_address, AMDGPU_VA_HOLE_START,
AMDGPU_VA_HOLE_END);
return -EINVAL;
}
args->va_address &= AMDGPU_VA_HOLE_MASK;
vm_size = adev->vm_manager.max_pfn * AMDGPU_GPU_PAGE_SIZE;
vm_size -= AMDGPU_VA_RESERVED_SIZE;
if (args->va_address + args->map_size > vm_size) {
dev_dbg(&dev->pdev->dev,
"va_address 0x%llx is in top reserved area 0x%llx\n",
args->va_address + args->map_size, vm_size);
return -EINVAL;
}
if ((args->flags & ~valid_flags) && (args->flags & ~prt_flags)) {
dev_dbg(&dev->pdev->dev, "invalid flags combination 0x%08X\n",
args->flags);
return -EINVAL;
}
switch (args->operation) {
case AMDGPU_VA_OP_MAP:
case AMDGPU_VA_OP_UNMAP:
case AMDGPU_VA_OP_CLEAR:
case AMDGPU_VA_OP_REPLACE:
break;
default:
dev_dbg(&dev->pdev->dev, "unsupported operation %d\n",
args->operation);
return -EINVAL;
}
INIT_LIST_HEAD(&list);
INIT_LIST_HEAD(&duplicates);
if ((args->operation != AMDGPU_VA_OP_CLEAR) &&
!(args->flags & AMDGPU_VM_PAGE_PRT)) {
gobj = drm_gem_object_lookup(filp, args->handle);
if (gobj == NULL)
return -ENOENT;
abo = gem_to_amdgpu_bo(gobj);
tv.bo = &abo->tbo;
tv.shared = !!(abo->flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID);
list_add(&tv.head, &list);
} else {
gobj = NULL;
abo = NULL;
}
amdgpu_vm_get_pd_bo(&fpriv->vm, &list, &vm_pd);
r = ttm_eu_reserve_buffers(&ticket, &list, true, &duplicates);
if (r)
goto error_unref;
if (abo) {
bo_va = amdgpu_vm_bo_find(&fpriv->vm, abo);
if (!bo_va) {
r = -ENOENT;
goto error_backoff;
}
} else if (args->operation != AMDGPU_VA_OP_CLEAR) {
bo_va = fpriv->prt_va;
} else {
bo_va = NULL;
}
switch (args->operation) {
case AMDGPU_VA_OP_MAP:
r = amdgpu_vm_alloc_pts(adev, bo_va->base.vm, args->va_address,
args->map_size);
if (r)
goto error_backoff;
va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags);
r = amdgpu_vm_bo_map(adev, bo_va, args->va_address,
args->offset_in_bo, args->map_size,
va_flags);
break;
case AMDGPU_VA_OP_UNMAP:
r = amdgpu_vm_bo_unmap(adev, bo_va, args->va_address);
break;
case AMDGPU_VA_OP_CLEAR:
r = amdgpu_vm_bo_clear_mappings(adev, &fpriv->vm,
args->va_address,
args->map_size);
break;
case AMDGPU_VA_OP_REPLACE:
r = amdgpu_vm_alloc_pts(adev, bo_va->base.vm, args->va_address,
args->map_size);
if (r)
goto error_backoff;
va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags);
r = amdgpu_vm_bo_replace_map(adev, bo_va, args->va_address,
args->offset_in_bo, args->map_size,
va_flags);
break;
default:
break;
}
if (!r && !(args->flags & AMDGPU_VM_DELAY_UPDATE) && !amdgpu_vm_debug)
amdgpu_gem_va_update_vm(adev, &fpriv->vm, bo_va,
args->operation);
error_backoff:
ttm_eu_backoff_reservation(&ticket, &list);
error_unref:
drm_gem_object_put_unlocked(gobj);
return r;
}
int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
struct amdgpu_device *adev = dev->dev_private;
struct drm_amdgpu_gem_op *args = data;
struct drm_gem_object *gobj;
struct amdgpu_bo *robj;
int r;
gobj = drm_gem_object_lookup(filp, args->handle);
if (gobj == NULL) {
return -ENOENT;
}
robj = gem_to_amdgpu_bo(gobj);
r = amdgpu_bo_reserve(robj, false);
if (unlikely(r))
goto out;
switch (args->op) {
case AMDGPU_GEM_OP_GET_GEM_CREATE_INFO: {
struct drm_amdgpu_gem_create_in info;
void __user *out = u64_to_user_ptr(args->value);
info.bo_size = robj->gem_base.size;
info.alignment = robj->tbo.mem.page_alignment << PAGE_SHIFT;
info.domains = robj->preferred_domains;
info.domain_flags = robj->flags;
amdgpu_bo_unreserve(robj);
if (copy_to_user(out, &info, sizeof(info)))
r = -EFAULT;
break;
}
case AMDGPU_GEM_OP_SET_PLACEMENT:
if (robj->prime_shared_count && (args->value & AMDGPU_GEM_DOMAIN_VRAM)) {
r = -EINVAL;
amdgpu_bo_unreserve(robj);
break;
}
if (amdgpu_ttm_tt_get_usermm(robj->tbo.ttm)) {
r = -EPERM;
amdgpu_bo_unreserve(robj);
break;
}
robj->preferred_domains = args->value & (AMDGPU_GEM_DOMAIN_VRAM |
AMDGPU_GEM_DOMAIN_GTT |
AMDGPU_GEM_DOMAIN_CPU);
robj->allowed_domains = robj->preferred_domains;
if (robj->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
robj->allowed_domains |= AMDGPU_GEM_DOMAIN_GTT;
if (robj->flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID)
amdgpu_vm_bo_invalidate(adev, robj, true);
amdgpu_bo_unreserve(robj);
break;
default:
amdgpu_bo_unreserve(robj);
r = -EINVAL;
}
out:
drm_gem_object_put_unlocked(gobj);
return r;
}
int amdgpu_mode_dumb_create(struct drm_file *file_priv,
struct drm_device *dev,
struct drm_mode_create_dumb *args)
{
struct amdgpu_device *adev = dev->dev_private;
struct drm_gem_object *gobj;
uint32_t handle;
u32 domain;
int r;
args->pitch = amdgpu_align_pitch(adev, args->width,
DIV_ROUND_UP(args->bpp, 8), 0);
args->size = (u64)args->pitch * args->height;
args->size = ALIGN(args->size, PAGE_SIZE);
domain = amdgpu_bo_get_preferred_pin_domain(adev,
amdgpu_display_supported_domains(adev));
r = amdgpu_gem_object_create(adev, args->size, 0, domain,
AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
ttm_bo_type_device, NULL, &gobj);
if (r)
return -ENOMEM;
r = drm_gem_handle_create(file_priv, gobj, &handle);
/* drop reference from allocate - handle holds it now */
drm_gem_object_put_unlocked(gobj);
if (r) {
return r;
}
args->handle = handle;
return 0;
}
#if defined(CONFIG_DEBUG_FS)
#define amdgpu_debugfs_gem_bo_print_flag(m, bo, flag) \
if (bo->flags & (AMDGPU_GEM_CREATE_ ## flag)) { \
seq_printf((m), " " #flag); \
}
static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
{
struct drm_gem_object *gobj = ptr;
struct amdgpu_bo *bo = gem_to_amdgpu_bo(gobj);
struct seq_file *m = data;
struct dma_buf_attachment *attachment;
struct dma_buf *dma_buf;
unsigned domain;
const char *placement;
unsigned pin_count;
domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
switch (domain) {
case AMDGPU_GEM_DOMAIN_VRAM:
placement = "VRAM";
break;
case AMDGPU_GEM_DOMAIN_GTT:
placement = " GTT";
break;
case AMDGPU_GEM_DOMAIN_CPU:
default:
placement = " CPU";
break;
}
seq_printf(m, "\t0x%08x: %12ld byte %s",
id, amdgpu_bo_size(bo), placement);
pin_count = READ_ONCE(bo->pin_count);
if (pin_count)
seq_printf(m, " pin count %d", pin_count);
dma_buf = READ_ONCE(bo->gem_base.dma_buf);
attachment = READ_ONCE(bo->gem_base.import_attach);
if (attachment)
seq_printf(m, " imported from %p", dma_buf);
else if (dma_buf)
seq_printf(m, " exported as %p", dma_buf);
amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_ACCESS_REQUIRED);
amdgpu_debugfs_gem_bo_print_flag(m, bo, NO_CPU_ACCESS);
amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_GTT_USWC);
amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CLEARED);
amdgpu_debugfs_gem_bo_print_flag(m, bo, SHADOW);
amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CONTIGUOUS);
amdgpu_debugfs_gem_bo_print_flag(m, bo, VM_ALWAYS_VALID);
amdgpu_debugfs_gem_bo_print_flag(m, bo, EXPLICIT_SYNC);
seq_printf(m, "\n");
return 0;
}
static int amdgpu_debugfs_gem_info(struct seq_file *m, void *data)
{
struct drm_info_node *node = (struct drm_info_node *)m->private;
struct drm_device *dev = node->minor->dev;
struct drm_file *file;
int r;
r = mutex_lock_interruptible(&dev->filelist_mutex);
if (r)
return r;
list_for_each_entry(file, &dev->filelist, lhead) {
struct task_struct *task;
/*
* Although we have a valid reference on file->pid, that does
* not guarantee that the task_struct who called get_pid() is
* still alive (e.g. get_pid(current) => fork() => exit()).
* Therefore, we need to protect this ->comm access using RCU.
*/
rcu_read_lock();
task = pid_task(file->pid, PIDTYPE_PID);
seq_printf(m, "pid %8d command %s:\n", pid_nr(file->pid),
task ? task->comm : "<unknown>");
rcu_read_unlock();
spin_lock(&file->table_lock);
idr_for_each(&file->object_idr, amdgpu_debugfs_gem_bo_info, m);
spin_unlock(&file->table_lock);
}
mutex_unlock(&dev->filelist_mutex);
return 0;
}
static const struct drm_info_list amdgpu_debugfs_gem_list[] = {
{"amdgpu_gem_info", &amdgpu_debugfs_gem_info, 0, NULL},
};
#endif
int amdgpu_debugfs_gem_init(struct amdgpu_device *adev)
{
#if defined(CONFIG_DEBUG_FS)
return amdgpu_debugfs_add_files(adev, amdgpu_debugfs_gem_list, 1);
#endif
return 0;
}