}
/**
+ * Check if can do 2M page
+ * @vgpu: target vgpu
+ * @entry: target pfn's gtt entry
+ *
* Return 1 if 2MB huge gtt shadowing is possilbe, 0 if miscondition,
* negtive if found err.
*/
vgpu_free_mm(mm);
return ERR_PTR(-ENOMEM);
}
- mm->ggtt_mm.last_partial_off = -1UL;
return mm;
}
invalidate_ppgtt_mm(mm);
} else {
vfree(mm->ggtt_mm.virtual_ggtt);
- mm->ggtt_mm.last_partial_off = -1UL;
}
vgpu_free_mm(mm);
/**
* intel_vgpu_pin_mm - increase the pin count of a vGPU mm object
- * @vgpu: a vGPU
+ * @mm: target vgpu mm
*
* This function is called when user wants to use a vGPU mm object. If this
* mm object hasn't been shadowed yet, the shadow will be populated at this
struct intel_gvt_gtt_entry e, m;
dma_addr_t dma_addr;
int ret;
+ struct intel_gvt_partial_pte *partial_pte, *pos, *n;
+ bool partial_update = false;
if (bytes != 4 && bytes != 8)
return -EINVAL;
if (!vgpu_gmadr_is_valid(vgpu, gma))
return 0;
- ggtt_get_guest_entry(ggtt_mm, &e, g_gtt_index);
-
+ e.type = GTT_TYPE_GGTT_PTE;
memcpy((void *)&e.val64 + (off & (info->gtt_entry_size - 1)), p_data,
bytes);
/* If ggtt entry size is 8 bytes, and it's split into two 4 bytes
- * write, we assume the two 4 bytes writes are consecutive.
- * Otherwise, we abort and report error
+ * write, save the first 4 bytes in a list and update virtual
+ * PTE. Only update shadow PTE when the second 4 bytes comes.
*/
if (bytes < info->gtt_entry_size) {
- if (ggtt_mm->ggtt_mm.last_partial_off == -1UL) {
- /* the first partial part*/
- ggtt_mm->ggtt_mm.last_partial_off = off;
- ggtt_mm->ggtt_mm.last_partial_data = e.val64;
- return 0;
- } else if ((g_gtt_index ==
- (ggtt_mm->ggtt_mm.last_partial_off >>
- info->gtt_entry_size_shift)) &&
- (off != ggtt_mm->ggtt_mm.last_partial_off)) {
- /* the second partial part */
-
- int last_off = ggtt_mm->ggtt_mm.last_partial_off &
- (info->gtt_entry_size - 1);
-
- memcpy((void *)&e.val64 + last_off,
- (void *)&ggtt_mm->ggtt_mm.last_partial_data +
- last_off, bytes);
-
- ggtt_mm->ggtt_mm.last_partial_off = -1UL;
- } else {
- int last_offset;
-
- gvt_vgpu_err("failed to populate guest ggtt entry: abnormal ggtt entry write sequence, last_partial_off=%lx, offset=%x, bytes=%d, ggtt entry size=%d\n",
- ggtt_mm->ggtt_mm.last_partial_off, off,
- bytes, info->gtt_entry_size);
-
- /* set host ggtt entry to scratch page and clear
- * virtual ggtt entry as not present for last
- * partially write offset
- */
- last_offset = ggtt_mm->ggtt_mm.last_partial_off &
- (~(info->gtt_entry_size - 1));
-
- ggtt_get_host_entry(ggtt_mm, &m, last_offset);
- ggtt_invalidate_pte(vgpu, &m);
- ops->set_pfn(&m, gvt->gtt.scratch_mfn);
- ops->clear_present(&m);
- ggtt_set_host_entry(ggtt_mm, &m, last_offset);
- ggtt_invalidate(gvt->dev_priv);
-
- ggtt_get_guest_entry(ggtt_mm, &e, last_offset);
- ops->clear_present(&e);
- ggtt_set_guest_entry(ggtt_mm, &e, last_offset);
-
- ggtt_mm->ggtt_mm.last_partial_off = off;
- ggtt_mm->ggtt_mm.last_partial_data = e.val64;
+ bool found = false;
+
+ list_for_each_entry_safe(pos, n,
+ &ggtt_mm->ggtt_mm.partial_pte_list, list) {
+ if (g_gtt_index == pos->offset >>
+ info->gtt_entry_size_shift) {
+ if (off != pos->offset) {
+ /* the second partial part*/
+ int last_off = pos->offset &
+ (info->gtt_entry_size - 1);
+
+ memcpy((void *)&e.val64 + last_off,
+ (void *)&pos->data + last_off,
+ bytes);
+
+ list_del(&pos->list);
+ kfree(pos);
+ found = true;
+ break;
+ }
+
+ /* update of the first partial part */
+ pos->data = e.val64;
+ ggtt_set_guest_entry(ggtt_mm, &e, g_gtt_index);
+ return 0;
+ }
+ }
- return 0;
+ if (!found) {
+ /* the first partial part */
+ partial_pte = kzalloc(sizeof(*partial_pte), GFP_KERNEL);
+ if (!partial_pte)
+ return -ENOMEM;
+ partial_pte->offset = off;
+ partial_pte->data = e.val64;
+ list_add_tail(&partial_pte->list,
+ &ggtt_mm->ggtt_mm.partial_pte_list);
+ partial_update = true;
}
}
- if (ops->test_present(&e)) {
+ if (!partial_update && (ops->test_present(&e))) {
gfn = ops->get_pfn(&e);
m = e;
} else
ops->set_pfn(&m, dma_addr >> PAGE_SHIFT);
} else {
- ggtt_get_host_entry(ggtt_mm, &m, g_gtt_index);
- ggtt_invalidate_pte(vgpu, &m);
ops->set_pfn(&m, gvt->gtt.scratch_mfn);
ops->clear_present(&m);
}
out:
+ ggtt_set_guest_entry(ggtt_mm, &e, g_gtt_index);
+
+ ggtt_get_host_entry(ggtt_mm, &e, g_gtt_index);
+ ggtt_invalidate_pte(vgpu, &e);
+
ggtt_set_host_entry(ggtt_mm, &m, g_gtt_index);
ggtt_invalidate(gvt->dev_priv);
- ggtt_set_guest_entry(ggtt_mm, &e, g_gtt_index);
return 0;
}
intel_vgpu_reset_ggtt(vgpu, false);
+ INIT_LIST_HEAD(>t->ggtt_mm->ggtt_mm.partial_pte_list);
+
return create_scratch_page_tree(vgpu);
}
static void intel_vgpu_destroy_ggtt_mm(struct intel_vgpu *vgpu)
{
+ struct intel_gvt_partial_pte *pos;
+
+ list_for_each_entry(pos,
+ &vgpu->gtt.ggtt_mm->ggtt_mm.partial_pte_list, list) {
+ gvt_dbg_mm("partial PTE update on hold 0x%lx : 0x%llx\n",
+ pos->offset, pos->data);
+ kfree(pos);
+ }
intel_vgpu_destroy_mm(vgpu->gtt.ggtt_mm);
vgpu->gtt.ggtt_mm = NULL;
}
/**
* intel_vgpu_find_ppgtt_mm - find a PPGTT mm object
* @vgpu: a vGPU
- * @page_table_level: PPGTT page table level
- * @root_entry: PPGTT page table root pointers
+ * @pdps: pdp root array
*
* This function is used to find a PPGTT mm object from mm object pool
*
#define _GVT_GTT_H_
#define I915_GTT_PAGE_SHIFT 12
-#define I915_GTT_PAGE_MASK (~(I915_GTT_PAGE_SIZE - 1))
struct intel_vgpu_mm;
#define GVT_RING_CTX_NR_PDPS GEN8_3LVL_PDPES
+ struct intel_gvt_partial_pte {
+ unsigned long offset;
+ u64 data;
+ struct list_head list;
+ };
+
struct intel_vgpu_mm {
enum intel_gvt_mm_type type;
struct intel_vgpu *vgpu;
} ppgtt_mm;
struct {
void *virtual_ggtt;
- unsigned long last_partial_off;
- u64 last_partial_data;
+ struct list_head partial_pte_list;
} ggtt_mm;
};
};
{
write_vreg(vgpu, offset, p_data, bytes);
- if (vgpu_vreg(vgpu, offset) & HSW_PWR_WELL_CTL_REQ(HSW_DISP_PW_GLOBAL))
+ if (vgpu_vreg(vgpu, offset) &
+ HSW_PWR_WELL_CTL_REQ(HSW_PW_CTL_IDX_GLOBAL))
vgpu_vreg(vgpu, offset) |=
- HSW_PWR_WELL_CTL_STATE(HSW_DISP_PW_GLOBAL);
+ HSW_PWR_WELL_CTL_STATE(HSW_PW_CTL_IDX_GLOBAL);
else
vgpu_vreg(vgpu, offset) &=
- ~HSW_PWR_WELL_CTL_STATE(HSW_DISP_PW_GLOBAL);
+ ~HSW_PWR_WELL_CTL_STATE(HSW_PW_CTL_IDX_GLOBAL);
return 0;
}
return 0;
}
- static int bxt_edp_psr_imr_iir_write(struct intel_vgpu *vgpu,
+ static int edp_psr_imr_iir_write(struct intel_vgpu *vgpu,
unsigned int offset, void *p_data, unsigned int bytes)
{
vgpu_vreg(vgpu, offset) = 0;
MMIO_F(PCH_GMBUS0, 4 * 4, 0, 0, 0, D_ALL, gmbus_mmio_read,
gmbus_mmio_write);
- MMIO_F(PCH_GPIOA, 6 * 4, F_UNALIGN, 0, 0, D_ALL, NULL, NULL);
+ MMIO_F(PCH_GPIO_BASE, 6 * 4, F_UNALIGN, 0, 0, D_ALL, NULL, NULL);
MMIO_F(_MMIO(0xe4f00), 0x28, 0, 0, 0, D_ALL, NULL, NULL);
MMIO_F(_MMIO(_PCH_DPB_AUX_CH_CTL), 6 * 4, 0, 0, 0, D_PRE_SKL, NULL,
MMIO_D(GEN6_RC6p_THRESHOLD, D_ALL);
MMIO_D(GEN6_RC6pp_THRESHOLD, D_ALL);
MMIO_D(GEN6_PMINTRMSK, D_ALL);
- /*
- * Use an arbitrary power well controlled by the PWR_WELL_CTL
- * register.
- */
- MMIO_DH(HSW_PWR_WELL_CTL_BIOS(HSW_DISP_PW_GLOBAL), D_BDW, NULL,
- power_well_ctl_mmio_write);
- MMIO_DH(HSW_PWR_WELL_CTL_DRIVER(HSW_DISP_PW_GLOBAL), D_BDW, NULL,
- power_well_ctl_mmio_write);
- MMIO_DH(HSW_PWR_WELL_CTL_KVMR, D_BDW, NULL, power_well_ctl_mmio_write);
- MMIO_DH(HSW_PWR_WELL_CTL_DEBUG(HSW_DISP_PW_GLOBAL), D_BDW, NULL,
- power_well_ctl_mmio_write);
+ MMIO_DH(HSW_PWR_WELL_CTL1, D_BDW, NULL, power_well_ctl_mmio_write);
+ MMIO_DH(HSW_PWR_WELL_CTL2, D_BDW, NULL, power_well_ctl_mmio_write);
+ MMIO_DH(HSW_PWR_WELL_CTL3, D_BDW, NULL, power_well_ctl_mmio_write);
+ MMIO_DH(HSW_PWR_WELL_CTL4, D_BDW, NULL, power_well_ctl_mmio_write);
MMIO_DH(HSW_PWR_WELL_CTL5, D_BDW, NULL, power_well_ctl_mmio_write);
MMIO_DH(HSW_PWR_WELL_CTL6, D_BDW, NULL, power_well_ctl_mmio_write);
MMIO_DFH(_MMIO(0x1a178), D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
MMIO_DFH(_MMIO(0x1a17c), D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
MMIO_DFH(_MMIO(0x2217c), D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL);
+
+ MMIO_DH(EDP_PSR_IMR, D_BDW_PLUS, NULL, edp_psr_imr_iir_write);
+ MMIO_DH(EDP_PSR_IIR, D_BDW_PLUS, NULL, edp_psr_imr_iir_write);
return 0;
}
MMIO_F(_MMIO(_DPD_AUX_CH_CTL), 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL,
dp_aux_ch_ctl_mmio_write);
- /*
- * Use an arbitrary power well controlled by the PWR_WELL_CTL
- * register.
- */
- MMIO_D(HSW_PWR_WELL_CTL_BIOS(SKL_DISP_PW_MISC_IO), D_SKL_PLUS);
- MMIO_DH(HSW_PWR_WELL_CTL_DRIVER(SKL_DISP_PW_MISC_IO), D_SKL_PLUS, NULL,
- skl_power_well_ctl_write);
+ MMIO_D(HSW_PWR_WELL_CTL1, D_SKL_PLUS);
+ MMIO_DH(HSW_PWR_WELL_CTL2, D_SKL_PLUS, NULL, skl_power_well_ctl_write);
MMIO_DH(DBUF_CTL, D_SKL_PLUS, NULL, gen9_dbuf_ctl_mmio_write);
MMIO_D(HSW_TVIDEO_DIP_GCP(TRANSCODER_B), D_BXT);
MMIO_D(HSW_TVIDEO_DIP_GCP(TRANSCODER_C), D_BXT);
- MMIO_DH(EDP_PSR_IMR, D_BXT, NULL, bxt_edp_psr_imr_iir_write);
- MMIO_DH(EDP_PSR_IIR, D_BXT, NULL, bxt_edp_psr_imr_iir_write);
-
MMIO_D(RC6_CTX_BASE, D_BXT);
MMIO_D(GEN8_PUSHBUS_CONTROL, D_BXT);
* @offset: register offset
* @pdata: data buffer
* @bytes: data length
+ * @is_read: read or write
*
* Returns:
* Zero on success, negative error code if failed.
#include "gvt.h"
#include "trace.h"
-/**
- * Defined in Intel Open Source PRM.
- * Ref: https://01.org/linuxgraphics/documentation/hardware-specification-prms
- */
-#define TRVATTL3PTRDW(i) _MMIO(0x4de0 + (i)*4)
-#define TRNULLDETCT _MMIO(0x4de8)
-#define TRINVTILEDETCT _MMIO(0x4dec)
-#define TRVADR _MMIO(0x4df0)
-#define TRTTE _MMIO(0x4df4)
-#define RING_EXCC(base) _MMIO((base) + 0x28)
-#define RING_GFX_MODE(base) _MMIO((base) + 0x29c)
-#define VF_GUARDBAND _MMIO(0x83a4)
-
#define GEN9_MOCS_SIZE 64
/* Raw offset is appened to each line for convenience. */
{RCS, GAMT_CHKN_BIT_REG, 0x0, false}, /* 0x4ab8 */
{RCS, GEN9_GAMT_ECO_REG_RW_IA, 0x0, false}, /* 0x4ab0 */
- {RCS, GEN9_CSFE_CHICKEN1_RCS, 0x0, false}, /* 0x20d4 */
+ {RCS, GEN9_CSFE_CHICKEN1_RCS, 0xffff, false}, /* 0x20d4 */
{RCS, GEN8_GARBCNTL, 0x0, false}, /* 0xb004 */
{RCS, GEN7_FF_THREAD_MODE, 0x0, false}, /* 0x20a0 */