intel: tools: split memory management out of aubinator

Signed-off-by: Lionel Landwerlin <[email protected]>
Reviewed-by: Rafael Antognolli <[email protected]>

1 files changed, 391 insertions, 0 deletions

diff --git a/src/intel/tools/aub_mem.c b/src/intel/tools/aub_mem.c
new file mode 100644
index 00000000000..92484272906
--- /dev/null
+++ b/src/intel/tools/aub_mem.c
@@ -0,0 +1,391 @@
+/*
+ * Copyright © 2016-2018 Intel Corporation
+ *
+ * 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 (including the next
+ * paragraph) 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 AUTHORS OR COPYRIGHT HOLDERS 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.
+ */
+
+#include <string.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/mman.h>
+
+#include "aub_mem.h"
+
+#ifndef HAVE_MEMFD_CREATE
+#include <sys/syscall.h>
+
+static inline int
+memfd_create(const char *name, unsigned int flags)
+{
+   return syscall(SYS_memfd_create, name, flags);
+}
+#endif
+
+struct bo_map {
+   struct list_head link;
+   struct gen_batch_decode_bo bo;
+   bool unmap_after_use;
+};
+
+struct ggtt_entry {
+   struct rb_node node;
+   uint64_t virt_addr;
+   uint64_t phys_addr;
+};
+
+struct phys_mem {
+   struct rb_node node;
+   uint64_t fd_offset;
+   uint64_t phys_addr;
+   uint8_t *data;
+};
+
+static void
+add_gtt_bo_map(struct aub_mem *mem, struct gen_batch_decode_bo bo, bool unmap_after_use)
+{
+   struct bo_map *m = calloc(1, sizeof(*m));
+
+   m->bo = bo;
+   m->unmap_after_use = unmap_after_use;
+   list_add(&m->link, &mem->maps);
+}
+
+void
+aub_mem_clear_bo_maps(struct aub_mem *mem)
+{
+   list_for_each_entry_safe(struct bo_map, i, &mem->maps, link) {
+      if (i->unmap_after_use)
+         munmap((void *)i->bo.map, i->bo.size);
+      list_del(&i->link);
+      free(i);
+   }
+}
+
+static inline struct ggtt_entry *
+ggtt_entry_next(struct ggtt_entry *entry)
+{
+   if (!entry)
+      return NULL;
+   struct rb_node *node = rb_node_next(&entry->node);
+   if (!node)
+      return NULL;
+   return rb_node_data(struct ggtt_entry, node, node);
+}
+
+static inline int
+cmp_uint64(uint64_t a, uint64_t b)
+{
+   if (a < b)
+      return -1;
+   if (a > b)
+      return 1;
+   return 0;
+}
+
+static inline int
+cmp_ggtt_entry(const struct rb_node *node, const void *addr)
+{
+   struct ggtt_entry *entry = rb_node_data(struct ggtt_entry, node, node);
+   return cmp_uint64(entry->virt_addr, *(const uint64_t *)addr);
+}
+
+static struct ggtt_entry *
+ensure_ggtt_entry(struct aub_mem *mem, uint64_t virt_addr)
+{
+   struct rb_node *node = rb_tree_search_sloppy(&mem->ggtt, &virt_addr,
+                                                cmp_ggtt_entry);
+   int cmp = 0;
+   if (!node || (cmp = cmp_ggtt_entry(node, &virt_addr))) {
+      struct ggtt_entry *new_entry = calloc(1, sizeof(*new_entry));
+      new_entry->virt_addr = virt_addr;
+      rb_tree_insert_at(&mem->ggtt, node, &new_entry->node, cmp > 0);
+      node = &new_entry->node;
+   }
+
+   return rb_node_data(struct ggtt_entry, node, node);
+}
+
+static struct ggtt_entry *
+search_ggtt_entry(struct aub_mem *mem, uint64_t virt_addr)
+{
+   virt_addr &= ~0xfff;
+
+   struct rb_node *node = rb_tree_search(&mem->ggtt, &virt_addr, cmp_ggtt_entry);
+
+   if (!node)
+      return NULL;
+
+   return rb_node_data(struct ggtt_entry, node, node);
+}
+
+static inline int
+cmp_phys_mem(const struct rb_node *node, const void *addr)
+{
+   struct phys_mem *mem = rb_node_data(struct phys_mem, node, node);
+   return cmp_uint64(mem->phys_addr, *(uint64_t *)addr);
+}
+
+static struct phys_mem *
+ensure_phys_mem(struct aub_mem *mem, uint64_t phys_addr)
+{
+   struct rb_node *node = rb_tree_search_sloppy(&mem->mem, &phys_addr, cmp_phys_mem);
+   int cmp = 0;
+   if (!node || (cmp = cmp_phys_mem(node, &phys_addr))) {
+      struct phys_mem *new_mem = calloc(1, sizeof(*new_mem));
+      new_mem->phys_addr = phys_addr;
+      new_mem->fd_offset = mem->mem_fd_len;
+
+      MAYBE_UNUSED int ftruncate_res = ftruncate(mem->mem_fd, mem->mem_fd_len += 4096);
+      assert(ftruncate_res == 0);
+
+      new_mem->data = mmap(NULL, 4096, PROT_READ | PROT_WRITE, MAP_SHARED,
+                           mem->mem_fd, new_mem->fd_offset);
+      assert(new_mem->data != MAP_FAILED);
+
+      rb_tree_insert_at(&mem->mem, node, &new_mem->node, cmp > 0);
+      node = &new_mem->node;
+   }
+
+   return rb_node_data(struct phys_mem, node, node);
+}
+
+static struct phys_mem *
+search_phys_mem(struct aub_mem *mem, uint64_t phys_addr)
+{
+   phys_addr &= ~0xfff;
+
+   struct rb_node *node = rb_tree_search(&mem->mem, &phys_addr, cmp_phys_mem);
+
+   if (!node)
+      return NULL;
+
+   return rb_node_data(struct phys_mem, node, node);
+}
+
+void
+aub_mem_local_write(void *_mem, uint64_t address,
+                    const void *data, uint32_t size)
+{
+   struct aub_mem *mem = _mem;
+   struct gen_batch_decode_bo bo = {
+      .map = data,
+      .addr = address,
+      .size = size,
+   };
+   add_gtt_bo_map(mem, bo, false);
+}
+
+void
+aub_mem_ggtt_entry_write(void *_mem, uint64_t address,
+                         const void *_data, uint32_t _size)
+{
+   struct aub_mem *mem = _mem;
+   uint64_t virt_addr = (address / sizeof(uint64_t)) << 12;
+   const uint64_t *data = _data;
+   size_t size = _size / sizeof(*data);
+   for (const uint64_t *entry = data;
+        entry < data + size;
+        entry++, virt_addr += 4096) {
+      struct ggtt_entry *pt = ensure_ggtt_entry(mem, virt_addr);
+      pt->phys_addr = *entry;
+   }
+}
+
+void
+aub_mem_phys_write(void *_mem, uint64_t phys_address,
+                   const void *data, uint32_t size)
+{
+   struct aub_mem *mem = _mem;
+   uint32_t to_write = size;
+   for (uint64_t page = phys_address & ~0xfff; page < phys_address + size; page += 4096) {
+      struct phys_mem *pmem = ensure_phys_mem(mem, page);
+      uint64_t offset = MAX2(page, phys_address) - page;
+      uint32_t size_this_page = MIN2(to_write, 4096 - offset);
+      to_write -= size_this_page;
+      memcpy(pmem->data + offset, data, size_this_page);
+      data = (const uint8_t *)data + size_this_page;
+   }
+}
+
+void
+aub_mem_ggtt_write(void *_mem, uint64_t virt_address,
+                   const void *data, uint32_t size)
+{
+   struct aub_mem *mem = _mem;
+   uint32_t to_write = size;
+   for (uint64_t page = virt_address & ~0xfff; page < virt_address + size; page += 4096) {
+      struct ggtt_entry *entry = search_ggtt_entry(mem, page);
+      assert(entry && entry->phys_addr & 0x1);
+
+      uint64_t offset = MAX2(page, virt_address) - page;
+      uint32_t size_this_page = MIN2(to_write, 4096 - offset);
+      to_write -= size_this_page;
+
+      uint64_t phys_page = entry->phys_addr & ~0xfff; /* Clear the validity bits. */
+      aub_mem_phys_write(mem, phys_page + offset, data, size_this_page);
+      data = (const uint8_t *)data + size_this_page;
+   }
+}
+
+struct gen_batch_decode_bo
+aub_mem_get_ggtt_bo(void *_mem, uint64_t address)
+{
+   struct aub_mem *mem = _mem;
+   struct gen_batch_decode_bo bo = {0};
+
+   list_for_each_entry(struct bo_map, i, &mem->maps, link)
+      if (i->bo.addr <= address && i->bo.addr + i->bo.size > address)
+         return i->bo;
+
+   address &= ~0xfff;
+
+   struct ggtt_entry *start =
+      (struct ggtt_entry *)rb_tree_search_sloppy(&mem->ggtt, &address,
+                                                 cmp_ggtt_entry);
+   if (start && start->virt_addr < address)
+      start = ggtt_entry_next(start);
+   if (!start)
+      return bo;
+
+   struct ggtt_entry *last = start;
+   for (struct ggtt_entry *i = ggtt_entry_next(last);
+        i && last->virt_addr + 4096 == i->virt_addr;
+        last = i, i = ggtt_entry_next(last))
+      ;
+
+   bo.addr = MIN2(address, start->virt_addr);
+   bo.size = last->virt_addr - bo.addr + 4096;
+   bo.map = mmap(NULL, bo.size, PROT_READ, MAP_SHARED | MAP_ANONYMOUS, -1, 0);
+   assert(bo.map != MAP_FAILED);
+
+   for (struct ggtt_entry *i = start;
+        i;
+        i = i == last ? NULL : ggtt_entry_next(i)) {
+      uint64_t phys_addr = i->phys_addr & ~0xfff;
+      struct phys_mem *phys_mem = search_phys_mem(mem, phys_addr);
+
+      if (!phys_mem)
+         continue;
+
+      uint32_t map_offset = i->virt_addr - address;
+      void *res = mmap((uint8_t *)bo.map + map_offset, 4096, PROT_READ,
+                       MAP_SHARED | MAP_FIXED, mem->mem_fd, phys_mem->fd_offset);
+      assert(res != MAP_FAILED);
+   }
+
+   add_gtt_bo_map(mem, bo, true);
+
+   return bo;
+}
+
+static struct phys_mem *
+ppgtt_walk(struct aub_mem *mem, uint64_t pml4, uint64_t address)
+{
+   uint64_t shift = 39;
+   uint64_t addr = pml4;
+   for (int level = 4; level > 0; level--) {
+      struct phys_mem *table = search_phys_mem(mem, addr);
+      if (!table)
+         return NULL;
+      int index = (address >> shift) & 0x1ff;
+      uint64_t entry = ((uint64_t *)table->data)[index];
+      if (!(entry & 1))
+         return NULL;
+      addr = entry & ~0xfff;
+      shift -= 9;
+   }
+   return search_phys_mem(mem, addr);
+}
+
+static bool
+ppgtt_mapped(struct aub_mem *mem, uint64_t pml4, uint64_t address)
+{
+   return ppgtt_walk(mem, pml4, address) != NULL;
+}
+
+struct gen_batch_decode_bo
+aub_mem_get_ppgtt_bo(void *_mem, uint64_t address)
+{
+   struct aub_mem *mem = _mem;
+   struct gen_batch_decode_bo bo = {0};
+
+   address &= ~0xfff;
+
+   if (!ppgtt_mapped(mem, mem->pml4, address))
+      return bo;
+
+   /* Map everything until the first gap since we don't know how much the
+    * decoder actually needs.
+    */
+   uint64_t end = address;
+   while (ppgtt_mapped(mem, mem->pml4, end))
+      end += 4096;
+
+   bo.addr = address;
+   bo.size = end - address;
+   bo.map = mmap(NULL, bo.size, PROT_READ, MAP_SHARED | MAP_ANONYMOUS, -1, 0);
+   assert(bo.map != MAP_FAILED);
+
+   for (uint64_t page = address; page < end; page += 4096) {
+      struct phys_mem *phys_mem = ppgtt_walk(mem, mem->pml4, page);
+
+      void *res = mmap((uint8_t *)bo.map + (page - bo.addr), 4096, PROT_READ,
+                       MAP_SHARED | MAP_FIXED, mem->mem_fd, phys_mem->fd_offset);
+      assert(res != MAP_FAILED);
+   }
+
+   add_gtt_bo_map(mem, bo, true);
+
+   return bo;
+}
+
+bool
+aub_mem_init(struct aub_mem *mem)
+{
+   memset(mem, 0, sizeof(*mem));
+
+   list_inithead(&mem->maps);
+
+   mem->mem_fd = memfd_create("phys memory", 0);
+
+   return mem->mem_fd != -1;
+}
+
+void
+aub_mem_fini(struct aub_mem *mem)
+{
+   if (mem->mem_fd == -1)
+      return;
+
+   aub_mem_clear_bo_maps(mem);
+
+
+   rb_tree_foreach_safe(struct ggtt_entry, entry, &mem->ggtt, node) {
+      rb_tree_remove(&mem->ggtt, &entry->node);
+      free(entry);
+   }
+   rb_tree_foreach_safe(struct phys_mem, entry, &mem->mem, node) {
+      rb_tree_remove(&mem->mem, &entry->node);
+      free(entry);
+   }
+
+   close(mem->mem_fd);
+   mem->mem_fd = -1;
+}