[U-Boot] [PATCH] malloc: Fix issue with calloc memory possibly being non-zero

Since we set #define MORECORE_CLEARS 1, the code assumes 'sbrk' always returns zero'd out memory. However since its possible that free() returns memory ack to sbrk() via malloc_trim we could possible get non-zero'd memory from sbrk() if it allocates back memory that was previously freed to it.
There are two possible solutions to this problem. 1. change #define MORECORE_CLEARS 0 2. memset to zero memory returned to sbrk.
We go with the second since the sbrk being called to free up memory should be pretty rare.
The following code problems an example test to show the issue. This test code was inserted right after the call to mem_malloc_init(). we could possible get non-zero'd memory from sbrk() if it allocates back memory that was previously freed to it.
There are two possible solutions to this problem. 1. change #define MORECORE_CLEARS 0 2. memset to zero memory returned to sbrk.
We go with the second since the sbrk being called to free up memory should be pretty rare.
The following code problems an example test to show the issue. This test code was inserted right after the call to mem_malloc_init().
... u8 *p2; int i;
printf("MALLOC TEST\n"); p1 = malloc(135176); printf("P1 = %p\n", p1); memset(p1, 0xab, 135176);
free(p1); p2 = calloc(4097, 1); printf("P2 = %p %p\n", p2, p2 + 4097);
for (i = 0; i < 4097; i++) { if (p2[i] != 0) printf("miscompare at byte %d got %x\n", i, p2[i]);
free(p2); printf("END MALLOC TEST\n\n"); ...
Signed-off-by: Kumar Gala galak@kernel.crashing.org --- common/dlmalloc.c | 7 +++++++ 1 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/common/dlmalloc.c b/common/dlmalloc.c index 4871f4b..e9bab09 100644 --- a/common/dlmalloc.c +++ b/common/dlmalloc.c @@ -1511,6 +1511,13 @@ void *sbrk(ptrdiff_t increment) ulong old = mem_malloc_brk; ulong new = old + increment;
+ /* + * if we are giving memory back make sure we clear it out since + * we set MORECORE_CLEARS to 1 + */ + if (increment < 0) + memset((void *)new, 0, -increment); + if ((new < mem_malloc_start) || (new > mem_malloc_end)) return (void *)MORECORE_FAILURE;

On Monday, November 15, 2010 15:51:14 Kumar Gala wrote:
returns memory ack to sbrk() via malloc_trim we could possible get
"ack" -> "back"
The following code problems an example test to show the issue.
i dont understand this sentence
we could possible get non-zero'd memory from sbrk() if it allocates back memory that was previously freed to it.
There are two possible solutions to this problem.
- change #define MORECORE_CLEARS 0
- memset to zero memory returned to sbrk.
We go with the second since the sbrk being called to free up memory should be pretty rare.
The following code problems an example test to show the issue. This test code was inserted right after the call to mem_malloc_init().
this looks like the changelog is partially duplicated -mike

Dear Kumar Gala,
In message 1289854274-7006-1-git-send-email-galak@kernel.crashing.org you wrote:
Since we set #define MORECORE_CLEARS 1, the code assumes 'sbrk' always returns zero'd out memory. However since its possible that free() returns memory ack to sbrk() via malloc_trim we could possible get non-zero'd memory from sbrk() if it allocates back memory that was previously freed to it.
I confirm that the test case works fine here, too. So:
Tested-by: Wolfgang Denk wd@denx.de
But the commit message needs serious rework before this can get applied.
Best regards,
Wolfgang Denk

On Nov 15, 2010, at 4:06 PM, Wolfgang Denk wrote:
Dear Kumar Gala,
In message 1289854274-7006-1-git-send-email-galak@kernel.crashing.org you wrote:
Since we set #define MORECORE_CLEARS 1, the code assumes 'sbrk' always returns zero'd out memory. However since its possible that free() returns memory ack to sbrk() via malloc_trim we could possible get non-zero'd memory from sbrk() if it allocates back memory that was previously freed to it.
I confirm that the test case works fine here, too. So:
Tested-by: Wolfgang Denk wd@denx.de
But the commit message needs serious rework before this can get applied.
Oops, will fix and repost.
- k
participants (3)
-
Kumar Gala
-
Mike Frysinger
-
Wolfgang Denk