[Bug malloc/25945] ASLR information leak via Safe-Linking and tcache or fastbin chunks.

["wangxuszcn at foxmail dot com" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=25945

--- Comment #11 from wangxu <wangxuszcn at foxmail dot com> ---
Created attachment 12540
  --> https://sourceware.org/bugzilla/attachment.cgi?id=12540&action=edit
patch2(tcache,fast,small,large,unsorted) and bench-malloc-xxx test result

(In reply to wangxu from comment #10)
> Created attachment 12538 [details]
> patch and bench-malloc-xxx test result
> 
> Hi Carlos
>   Sorry for late reply.
>   I just clear other pointer from tcache and fastbin(without small,large).
> 
> 1. bench-malloc-simple-[8-64] looks like better performance than before.
> 2. bench-malloc-simple-[128-4096] and bench-malloc-thread-xx are almost the
> same.
> 
> Attachment is the results of two rounds testing of bench-malloc-simple and
> bench-malloc-thread.

Hi Carlos

If clear pointer from tcache,fast-small-large-unsorted bin.
1. bench-malloc-simple-[8-64] look better than before.
2. bench-malloc-simple-[128-1024] are almost the same.
3. bench-malloc-simple-[2048-4096] are slightly bad.
3. bench-malloc-thread-[6-32] need more than about 10% max_rss than before.
Attachment file is "patch2(tcache,fast,small,large,unsorted) and
bench-malloc-xxx test result.zip"

Taking into account the severity of the issue, the performance of the two
patches, I suggest consider only tcache and fastbin.

Thanks for any other suggestions.

-- 
You are receiving this mail because:
You are on the CC list for the bug.