linux-kernel - [PATCH 0/5] PERF fixes

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-Id: <20120325.162801.524272241373730265.davem@davemloft.net>
Date:	Sun, 25 Mar 2012 16:28:01 -0400 (EDT)
From:	David Miller <davem@...emloft.net>
To:	acme@...stprotocols.net
CC:	linux-kernel@...r.kernel.org
Subject: [PATCH 0/5] PERF fixes

This fixes the histogram out-of-date map entry issue as well
as other problems I hit or discovered along the way.

To be honest the histogram entry map pointer is completely
superfluous.

When histogram entries are being used when processing a perf
event, we have the computed and most uptodate map already
in the addr_location that has been built, so we can just
pass that map pointer down into perf_top__record_precise_ip()
and similar.

The other cases, when annotating, are also superfluous.  We don't need
the map file at all to compute the address to pass into addr2line and
objdump.  We have exactly two cases:

1) Dynamic object, use the raw unchanged address

2) ET_EXEC or .gnu.prelink_undo section, we can read the ELF file
   PT_LOAD entries or similar to adjust the address properly.

So neither case needs the map pointer at all.  Then all we're
left with are perf internal debugging printf's and those can
be simply removed.

Signed-off-by: David S. Miller <davem@...emloft.net>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/