<!DOCTYPE html><html><head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body>
    <p>Hi Joonkyo,<br>
      Sorry for the delay. <br>
      Yes, sure, I reproduced issue 2 (null-ptr-deref in amdgpu) and I
      will provide the fix soon.<br>
      However, issue No. 3 is no longer reproducible if the recent patch
      inflight is applied which fixes issue No 1.</p>
    <p>Do you see the same behavior?<br>
    </p>
    <p>Thanks in advance, Vitaly<br>
    </p>
    <div class="moz-cite-prefix">On 2024-03-07 20:18, Joonkyo Jung
      wrote:<br>
    </div>
    <blockquote type="cite" cite="mid:CAKc8oVU8C-MRVa640603R2Gi7BRZ5P8ZN_uJoruDzR_A8xg71Q@mail.gmail.com">
      
      <div>
        <div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">Hello, </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">thank
            you for patching the first bug we have sent!</div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)"><br>
          </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">Just
            a quick touch base with you, to ask if there has been any
            update on our other two bugs.</div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">They
            were each sent with emails titled </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">"Reporting
            a slab-use-after-free in amdgpu" (this one)</div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">"Reporting
            a null-ptr-deref in amdgpu". </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)"><br>
          </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">Thank
            you! </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)"><br>
          </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">Best, </div>
          <div dir="auto" style="font-family:'-apple-system','helveticaneue';font-size:15px;font-style:normal;font-weight:400;letter-spacing:normal;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;background-color:rgba(0,0,0,0);border-color:rgb(0,0,0);color:rgb(0,0,0)">Joonkyo</div>
        </div>
        <br>
      </div>
      <div><br>
        <div class="gmail_quote">
          <div dir="ltr" class="gmail_attr">2024년 2월 16일 (금) 오후 6:22,
            Joonkyo Jung <<a href="mailto:joonkyoj@yonsei.ac.kr" moz-do-not-send="true" class="moz-txt-link-freetext">joonkyoj@yonsei.ac.kr</a>>님이
            작성:<br>
          </div>
          <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div dir="ltr">Hello,<br>
              <br>
              We would like to report a slab-use-after-free bug in the
              AMDGPU DRM driver in the linux kernel v6.8-rc4 that we
              found with our customized Syzkaller.<br>
              The bug can be triggered by sending two ioctls to the
              AMDGPU DRM driver in succession.<br>
              <br>
              In amdgpu_bo_move, struct ttm_resource *old_mem =
              bo->resource is assigned.<br>
              As you can see on the alloc & free stack calls, on the
              same function amdgpu_bo_move,<br>
              amdgpu_move_blit in the end frees bo->resource at
              ttm_bo_move_accel_cleanup with ttm_bo_wait_free_node(bo,
              man->use_tt).<br>
              But amdgpu_bo_move continues after that, reaching
              trace_amdgpu_bo_move(abo, new_mem->mem_type,
              old_mem->mem_type) at the end, causing the
              use-after-free bug.<br>
              <br>
              Steps to reproduce are as below.<br>
              union drm_amdgpu_gem_create *arg1;<br>
              <br>
              arg1 = malloc(sizeof(union drm_amdgpu_gem_create));<br>
              arg1->in.bo_size = 0x8;<br>
              arg1->in.alignment = 0x0;<br>
              arg1->in.domains = 0x4;<br>
              arg1->in.domain_flags = 0x9;<br>
              ioctl(fd, 0xc0206440, arg1);<br>
              <br>
              arg1->in.bo_size = 0x7fffffff;<br>
              arg1->in.alignment = 0x0;<br>
              arg1->in.domains = 0x4;<br>
              arg1->in.domain_flags = 0x9;<br>
              ioctl(fd, 0xc0206440, arg1);<br>
              <br>
              The KASAN report is as follows:<br>
==================================================================<br>
              BUG: KASAN: slab-use-after-free in
              amdgpu_bo_move+0x1479/0x1550<br>
              Read of size 4 at addr ffff88800f5bee80 by task
              syz-executor/219<br>
              Call Trace:<br>
               <TASK><br>
               amdgpu_bo_move+0x1479/0x1550<br>
               ttm_bo_handle_move_mem+0x4d0/0x700<br>
               ttm_mem_evict_first+0x945/0x1230<br>
               ttm_bo_mem_space+0x6c7/0x940<br>
               ttm_bo_validate+0x286/0x650<br>
               ttm_bo_init_reserved+0x34c/0x490<br>
               amdgpu_bo_create+0x94b/0x1610<br>
               amdgpu_bo_create_user+0xa3/0x130<br>
               amdgpu_gem_create_ioctl+0x4bc/0xc10<br>
               drm_ioctl_kernel+0x300/0x410<br>
               drm_ioctl+0x648/0xb30<br>
               amdgpu_drm_ioctl+0xc8/0x160<br>
               </TASK><br>
              <br>
              Allocated by task 219:<br>
               kmalloc_trace+0x211/0x390<br>
               amdgpu_vram_mgr_new+0x1d6/0xbe0<br>
               ttm_resource_alloc+0xfd/0x1e0<br>
               ttm_bo_mem_space+0x255/0x940<br>
               ttm_bo_validate+0x286/0x650<br>
               ttm_bo_init_reserved+0x34c/0x490<br>
               amdgpu_bo_create+0x94b/0x1610<br>
               amdgpu_bo_create_user+0xa3/0x130<br>
               amdgpu_gem_create_ioctl+0x4bc/0xc10<br>
               drm_ioctl_kernel+0x300/0x410<br>
               drm_ioctl+0x648/0xb30<br>
               amdgpu_drm_ioctl+0xc8/0x160<br>
              <br>
              Freed by task 219:<br>
               kfree+0x111/0x2d0<br>
               ttm_resource_free+0x17e/0x1e0<br>
               ttm_bo_move_accel_cleanup+0x77e/0x9b0<br>
               amdgpu_move_blit+0x3db/0x670<br>
               amdgpu_bo_move+0xfa2/0x1550<br>
               ttm_bo_handle_move_mem+0x4d0/0x700<br>
               ttm_mem_evict_first+0x945/0x1230<br>
               ttm_bo_mem_space+0x6c7/0x940<br>
               ttm_bo_validate+0x286/0x650<br>
               ttm_bo_init_reserved+0x34c/0x490<br>
               amdgpu_bo_create+0x94b/0x1610<br>
               amdgpu_bo_create_user+0xa3/0x130<br>
               amdgpu_gem_create_ioctl+0x4bc/0xc10<br>
               drm_ioctl_kernel+0x300/0x410<br>
               drm_ioctl+0x648/0xb30<br>
               amdgpu_drm_ioctl+0xc8/0x160<br>
              <br>
              The buggy address belongs to the object at
              ffff88800f5bee70<br>
               which belongs to the cache kmalloc-96 of size 96<br>
              The buggy address is located 16 bytes inside of<br>
               freed 96-byte region [ffff88800f5bee70, ffff88800f5beed0)<br>
              <br>
              Should you need any more information, please do not
              hesitate to contact us.<br>
              <br>
              Best regards,<br>
              Joonkyo Jung<br>
            </div>
          </blockquote>
        </div>
      </div>
    </blockquote>
  </body>
</html>