-
Type:
Bug
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: None
-
Component/s: None
-
Labels:
test_tiered08.py failed in an unrelated patch build:
Here's the stack trace:
(gdb) bt #0 0x00007fc60f96369c in __wt_block_discard (session=session@entry=0x55ba9cad1cd0, block=block@entry=0x55ba9cb69800, added_size=4096) at ../src/block/block_write.c:64 #1 0x00007fc60f9612d4 in __wt_bm_read (bm=0x55ba9e2ee8c0, session=0x55ba9cad1cd0, buf=0x7ffda74656d0, addr=<optimized out>, addr_size=<optimized out>) at ../src/block/block_read.c:44 #2 0x00007fc60f969066 in __bm_read (bm=<optimized out>, session=<optimized out>, buf=<optimized out>, addr=<optimized out>, addr_size=<optimized out>) at ../src/block_cache/block_mgr.c:486 #3 0x00007fc60f9672d0 in __wt_blkcache_read (session=session@entry=0x55ba9cad1cd0, buf=buf@entry=0x7ffda74656d0, addr=addr@entry=0x7ffda7465701 "\300\206\201\344\a\241 b\001\206\247\375\177", addr_size=10) at ../src/block_cache/block_io.c:111 #4 0x00007fc60f9c1677 in __page_read (session=session@entry=0x55ba9cad1cd0, ref=ref@entry=0x55ba9e1b7360, flags=flags@entry=128) at ../src/btree/bt_read.c:183 #5 0x00007fc60f9c24aa in __wt_page_in_func (session=session@entry=0x55ba9cad1cd0, ref=ref@entry=0x55ba9e1b7360, flags=flags@entry=128, func=func@entry=0x7fc60fbab9c0 <__PRETTY_FUNCTION__.48765> "__wt_row_search", line=line@entry=438) at ../src/btree/bt_read.c:307 #6 0x00007fc60fa0bab0 in __wt_page_swap_func (line=438, func=0x7fc60fbab9c0 <__PRETTY_FUNCTION__.48765> "__wt_row_search", flags=128, want=0x55ba9e1b7360, held=0x55ba9e7e6140, session=0x55ba9cad1cd0) at ../src/include/btree_inline.h:2150 #7 __wt_row_search (cbt=cbt@entry=0x55ba9d6ad600, srch_key=srch_key@entry=0x55ba9d6ad718, insert=insert@entry=true, leaf=leaf@entry=0x0, leaf_safe=leaf_safe@entry=false, leaf_foundp=leaf_foundp@entry=0x0) at ../src/btree/row_srch.c:438 #8 0x00007fc60f98c047 in __cursor_row_search (leaf_foundp=0x0, leaf=0x0, insert=true, cbt=0x55ba9d6ad600) at ../src/btree/bt_cursor.c:501 #9 __wt_btcur_insert (cbt=cbt@entry=0x55ba9d6ad600) at ../src/btree/bt_cursor.c:1161 #10 0x00007fc60fa4b7df in __curfile_insert (cursor=0x55ba9d6ad600) at ../src/cursor/cur_file.c:384 #11 0x00007fc610196306 in _wrap_Cursor_insert (self=<optimized out>, args=<optimized out>) at lang/python/wiredtigerPYTHON_wrap.c:4083
Looking at the cause of the segfault:
(gdb) frame 0
#0 0x00007fc60f96369c in __wt_block_discard (session=session@entry=0x55ba9cad1cd0,
block=block@entry=0x55ba9cb69800, added_size=4096) at ../src/block/block_write.c:64
64 handle = block->fh->handle;
(gdb) print block
$4 = (WT_BLOCK *) 0x55ba9cb69800
(gdb) print block->fh
$5 = (WT_FH *) 0xabababababababab
(gdb) print *block
$6 = {name = 0x0, objectid = 2880154539, ref = 2880154539, q = {tqe_next = 0xabababababababab,
tqe_prev = 0xabababababababab}, hashq = {tqe_next = 0xabababababababab,
...
In other words we are trying to use block after it has been freed. This appears to be a race with whoever freed it. This function was called from __wt_bm_read() here:
/* Read the block. */ __wt_capacity_throttle(session, size, WT_THROTTLE_READ); WT_RET(__wt_block_read_off(session, block, buf, objectid, offset, size, checksum)); /* Optionally discard blocks from the system's buffer cache. */ WT_RET(__wt_block_discard(session, block, (size_t)size));
In other words, by the time we get into __wt_block_discard(), block has been freed. But we successfully completed the previous call into __wt_block_read_off() which used block. So some other thread must have freed it in the interim.
Note that this patch includes WT-7833.
- depends on
-
WT-10734 Move cache of block handles to block manager layer
-
- Closed
-
-
WT-10462 Change block manager to track all of the WT_BLOCKs for a tree
-
- Closed
-
- is depended on by
-
-
- Closed
-
- is duplicated by
-
-
- Closed
-
- is related to
-
-
- Closed
-
-
-
- Closed
-
- related to
-
-
- Open
-