linux-kernel - Re: [bug report] perf jit: add source line info support

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>] [<thread-prev] [day] [month] [year] [list]

Date:   Fri, 9 Dec 2016 21:23:28 -0800
From:   Stephane Eranian <eranian@...gle.com>
To:     Dan Carpenter <dan.carpenter@...cle.com>
Cc:     Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [bug report] perf jit: add source line info support

On Fri, Dec 9, 2016 at 9:16 PM, Stephane Eranian <eranian@...gle.com> wrote:
> Hi Dan,
>
> On Wed, Nov 30, 2016 at 10:48 AM, Dan Carpenter
> <dan.carpenter@...cle.com> wrote:
>> Hello Stephane Eranian,
>>
>> The patch 598b7c6919c7: "perf jit: add source line info support" from
>> Nov 30, 2015, leads to the following static checker warning:
>>
>>         ./tools/perf/util/genelf_debug.c:211 emit_signed_LEB128()
>>         warn: potential left shift more than type allows '57'
>>
>> ./tools/perf/util/genelf_debug.c
>>    202  static void emit_signed_LEB128(struct buffer_ext *be, long data)
>>    203  {
>>    204          int more = 1;
>>    205          int negative = data < 0;
>>    206          int size = sizeof(long) * CHAR_BIT;
>>    207          while (more) {
>>    208                  ubyte cur = data & 0x7F;
>>    209                  data >>= 7;
>>    210                  if (negative)
>>    211                          data |= - (1 << (size - 7));
>>                                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> This is a no-op on 64 bit systems.  I suspect it's not intentional?
>>
> I suspect size should be: size = sizeof(int) * CHAR_BIT;
> I will test that.
Sorry, I think the issue is on the 1 shift, should be 1UL instead.
>
>>    212                  if ((data == 0 && !(cur & 0x40)) ||
>>    213                      (data == -1l && (cur & 0x40)))
>>    214                          more = 0;
>>    215                  else
>>    216                          cur |= 0x80;
>>    217                  buffer_ext_add(be, &cur, 1);
>>    218          }
>>    219  }
>>
>> regards,
>> dan carpenter