Re: BUG #16527: Valgrind detects an invalid read in brin_revmap_data with non-index page - Mailing list pgsql-bugs

On Sat, Jul 04, 2020 at 04:00:00PM +0000, PG Bug reporting form wrote:
>The following bug has been logged on the website:
>
>Bug reference:      16527
>Logged by:          Alexander Lakhin
>Email address:      exclusion@gmail.com
>PostgreSQL version: 13beta2
>Operating system:   Ubuntu 20.04
>Description:
>
>Running the following query (based on contrib/pageinspect/sql/brin.sql)
>under valgrind:
>CREATE EXTENSION pageinspect;
>CREATE TABLE test1 (a int, b text);
>INSERT INTO test1 VALUES (1, 'one');
>SELECT * FROM brin_revmap_data(get_raw_page('test1', 0));
>
>leads to a memory access error:
>==00:00:00:12.518 934833== Invalid read of size 2
>==00:00:00:12.518 934833==    at 0x4865AE1: verify_brin_page
>(brinfuncs.c:107)
>==00:00:00:12.518 934833==    by 0x486674E: brin_revmap_data
>(brinfuncs.c:386)
>==00:00:00:12.518 934833==    by 0x3C9656: ExecMakeTableFunctionResult
>(execSRF.c:234)
>==00:00:00:12.518 934833==    by 0x3DB7D4: FunctionNext
>(nodeFunctionscan.c:95)
>==00:00:00:12.518 934833==    by 0x3CA059: ExecScanFetch (execScan.c:133)
>==00:00:00:12.518 934833==    by 0x3CA0F4: ExecScan (execScan.c:182)
>==00:00:00:12.518 934833==    by 0x3DB6DF: ExecFunctionScan
>(nodeFunctionscan.c:270)
>==00:00:00:12.518 934833==    by 0x3C70B2: ExecProcNodeFirst
>(execProcnode.c:450)
>==00:00:00:12.518 934833==    by 0x3BFDD3: ExecProcNode (executor.h:245)
>==00:00:00:12.518 934833==    by 0x3BFDD3: ExecutePlan (execMain.c:1646)
>==00:00:00:12.518 934833==    by 0x3BFFB3: standard_ExecutorRun
>(execMain.c:364)
>==00:00:00:12.518 934833==    by 0x3C007F: ExecutorRun (execMain.c:308)
>==00:00:00:12.518 934833==    by 0x55F21F: PortalRunSelect (pquery.c:912)
>==00:00:00:12.518 934833==  Address 0xe69cc0a is 2 bytes after a block of
>size 8,264 alloc'd
>==00:00:00:12.518 934833==    at 0x483B7F3: malloc (in
>/usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so)
>==00:00:00:12.518 934833==    by 0x6A94CE: AllocSetAlloc (aset.c:739)
>==00:00:00:12.518 934833==    by 0x6B2AA7: palloc (mcxt.c:963)
>==00:00:00:12.518 934833==    by 0x486B838: get_raw_page_internal
>(rawpage.c:154)
>==00:00:00:12.518 934833==    by 0x486BC35: get_raw_page (rawpage.c:62)
>==00:00:00:12.518 934833==    by 0x3BACBE: ExecInterpExpr
>(execExprInterp.c:699)
>==00:00:00:12.518 934833==    by 0x3B7A64: ExecInterpExprStillValid
>(execExprInterp.c:1802)
>==00:00:00:12.518 934833==    by 0x3C8C3B: ExecEvalExpr (executor.h:294)
>==00:00:00:12.518 934833==    by 0x3C8C3B: ExecEvalFuncArgs
>(execSRF.c:836)
>==00:00:00:12.518 934833==    by 0x3C95C8: ExecMakeTableFunctionResult
>(execSRF.c:181)
>==00:00:00:12.518 934833==    by 0x3DB7D4: FunctionNext
>(nodeFunctionscan.c:95)
>==00:00:00:12.518 934833==    by 0x3CA059: ExecScanFetch (execScan.c:133)
>==00:00:00:12.518 934833==    by 0x3CA0F4: ExecScan (execScan.c:182)
>==00:00:00:12.518 934833==
>{
>   <insert_a_suppression_name_here>
>   Memcheck:Addr2
>   fun:verify_brin_page
>   fun:brin_revmap_data
>   fun:ExecMakeTableFunctionResult
>   fun:FunctionNext
>   fun:ExecScanFetch
>   fun:ExecScan
>   fun:ExecFunctionScan
>   fun:ExecProcNodeFirst
>   fun:ExecProcNode
>   fun:ExecutePlan
>   fun:standard_ExecutorRun
>   fun:ExecutorRun
>   fun:PortalRunSelect
>}
>==00:00:00:12.519 934833== Invalid read of size 2
>==00:00:00:12.519 934833==    at 0x4865C07: verify_brin_page
>(brinfuncs.c:108)
>==00:00:00:12.519 934833==    by 0x486674E: brin_revmap_data
>(brinfuncs.c:386)
>==00:00:00:12.519 934833==    by 0x3C9656: ExecMakeTableFunctionResult
>(execSRF.c:234)
>==00:00:00:12.519 934833==    by 0x3DB7D4: FunctionNext
>(nodeFunctionscan.c:95)
>==00:00:00:12.519 934833==    by 0x3CA059: ExecScanFetch (execScan.c:133)
>==00:00:00:12.519 934833==    by 0x3CA0F4: ExecScan (execScan.c:182)
>==00:00:00:12.519 934833==    by 0x3DB6DF: ExecFunctionScan
>(nodeFunctionscan.c:270)
>==00:00:00:12.519 934833==    by 0x3C70B2: ExecProcNodeFirst
>(execProcnode.c:450)
>==00:00:00:12.519 934833==    by 0x3BFDD3: ExecProcNode (executor.h:245)
>==00:00:00:12.519 934833==    by 0x3BFDD3: ExecutePlan (execMain.c:1646)
>==00:00:00:12.519 934833==    by 0x3BFFB3: standard_ExecutorRun
>(execMain.c:364)
>==00:00:00:12.519 934833==    by 0x3C007F: ExecutorRun (execMain.c:308)
>==00:00:00:12.519 934833==    by 0x55F21F: PortalRunSelect (pquery.c:912)
>==00:00:00:12.519 934833==  Address 0xe69cc0a is 2 bytes after a block of
>size 8,264 alloc'd
>==00:00:00:12.519 934833==    at 0x483B7F3: malloc (in
>/usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so)
>==00:00:00:12.519 934833==    by 0x6A94CE: AllocSetAlloc (aset.c:739)
>==00:00:00:12.519 934833==    by 0x6B2AA7: palloc (mcxt.c:963)
>==00:00:00:12.519 934833==    by 0x486B838: get_raw_page_internal
>(rawpage.c:154)
>==00:00:00:12.519 934833==    by 0x486BC35: get_raw_page (rawpage.c:62)
>==00:00:00:12.519 934833==    by 0x3BACBE: ExecInterpExpr
>(execExprInterp.c:699)
>==00:00:00:12.519 934833==    by 0x3B7A64: ExecInterpExprStillValid
>(execExprInterp.c:1802)
>==00:00:00:12.519 934833==    by 0x3C8C3B: ExecEvalExpr (executor.h:294)
>==00:00:00:12.519 934833==    by 0x3C8C3B: ExecEvalFuncArgs
>(execSRF.c:836)
>==00:00:00:12.519 934833==    by 0x3C95C8: ExecMakeTableFunctionResult
>(execSRF.c:181)
>==00:00:00:12.519 934833==    by 0x3DB7D4: FunctionNext
>(nodeFunctionscan.c:95)
>==00:00:00:12.519 934833==    by 0x3CA059: ExecScanFetch (execScan.c:133)
>==00:00:00:12.519 934833==    by 0x3CA0F4: ExecScan (execScan.c:182)
>==00:00:00:12.519 934833==
>{
>   <insert_a_suppression_name_here>
>   Memcheck:Addr2
>   fun:verify_brin_page
>   fun:brin_revmap_data
>   fun:ExecMakeTableFunctionResult
>   fun:FunctionNext
>   fun:ExecScanFetch
>   fun:ExecScan
>   fun:ExecFunctionScan
>   fun:ExecProcNodeFirst
>   fun:ExecProcNode
>   fun:ExecutePlan
>   fun:standard_ExecutorRun
>   fun:ExecutorRun
>   fun:PortalRunSelect
>}
>2020-07-04 17:57:55.915 MSK [934833] ERROR:  page is not a BRIN page of type
>"revmap"
>2020-07-04 17:57:55.915 MSK [934833] DETAIL:  Expected special type
>0000f092, got 00007f7f.

Hmmm, the 7f7f kinda seems like the pattern we use for randomizing
allocated/freed memory. So I thought maybe we're not initializing the
memory properly, or maybe freeing it too early. But I was getting
different patterns, and the reality is way simpler:


test=# SELECT * FROM page_header(get_raw_page('test1', 0));
     lsn    | checksum | flags | lower | upper | special | pagesize | version | prune_xid
-----------+----------+-------+-------+-------+---------+----------+---------+-----------
  0/15BBE80 |        0 |     4 |    28 |  8160 |    8192 |     8192 |       4 |         0
(1 row)

So the page actually does not have any special part, which is where the
type is supposed to be stored. So the BrinPageType probably ends up
reading whatever is immediately after the page. Interesting.

It might be worth adding an assert to check the PageGetSpecialPointer
result is actually within the page.


regards

-- 
Tomas Vondra                  http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services