Updating stack for 22450

Updating stack for 22450

TBR=dank@chromium.org
BUG=22450
TEST=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=53720

[stuartmorgan, 2010-07-27 00:18:34]

http://codereview.chromium.org/2832084/diff/2001/3001
File tools/valgrind/memcheck/suppressions.txt (right):

http://codereview.chromium.org/2832084/diff/2001/3001#newcode1498
tools/valgrind/memcheck/suppressions.txt:1498:
fun:_ZN3net64_GLOBAL__N_net_socket_client_socket_factory.cc_00000000_EA9FDEC926DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE
Anonymous functions tend to have fragile signatures. I would recommend:
  
fun:*DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE

[oshima, 2010-07-27 05:28:35]

Thank you for the tip

On Mon, Jul 26, 2010 at 5:18 PM, <stuartmorgan@chromium.org> wrote:

>
> http://codereview.chromium.org/2832084/diff/2001/3001
> File tools/valgrind/memcheck/suppressions.txt (right):
>
> http://codereview.chromium.org/2832084/diff/2001/3001#newcode1498
> tools/valgrind/memcheck/suppressions.txt:1498:
>
>
>
fun:_ZN3net64_GLOBAL__N_net_socket_client_socket_factory.cc_00000000_EA9FDEC926DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE
> Anonymous functions tend to have fragile signatures. I would recommend:
>
>
>
fun:*DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE


I'll fix this soon. Thank you for the tip.

- oshima

>
>
> http://codereview.chromium.org/2832084/show
>

[dank, 2010-07-27 15:36:56]

What Stuart said :-)

[oshima, 2010-07-28 06:05:28]

On Mon, Jul 26, 2010 at 5:18 PM, <stuartmorgan@chromium.org> wrote:

>
> http://codereview.chromium.org/2832084/diff/2001/3001
> File tools/valgrind/memcheck/suppressions.txt (right):
>
> http://codereview.chromium.org/2832084/diff/2001/3001#newcode1498
> tools/valgrind/memcheck/suppressions.txt:1498:
>
>
>
fun:_ZN3net64_GLOBAL__N_net_socket_client_socket_factory.cc_00000000_EA9FDEC926DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE
> Anonymous functions tend to have fragile signatures. I would recommend:
>
>
>
fun:*DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE


Hi,

I'm looking at suppressions.txt to fix other suppressions that uses
anonymous functions, and found that some suppressions are using the
following pattern.

fun:_ZN3net64_GLOBAL__N_net_socket_client_socket_factory.cc_00000000_*DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE

Which one do you think is better?

- oshima


>
>
> http://codereview.chromium.org/2832084/show
>

[Alexander Potapenko, 2010-07-28 06:22:13]

> 
> Hi,
> 
> I'm looking at suppressions.txt to fix other suppressions that uses
> anonymous functions, and found that some suppressions are using the
> following pattern.
> 
>
fun:_ZN3net64_GLOBAL__N_net_socket_client_socket_factory.cc_00000000_*DefaultClientSocketFactory21CreateTCPClientSocketERKNS_11AddressListEPNS_6NetLogE
> 
> Which one do you think is better?
> 
> - oshima
> 
> 

As this pattern is less strict, it's ok to use it until something breaks down
(it will if the file path changes). But as
"_GLOBAL__N_net_socket_client_socket_factory.cc_00000000_EA9FDEC9" is an unique
anonymous namespace name, we don't actually need it.
I suggest to take Stuart's version because it's more readable.

There are also the following patterns in suppressions.txt:
   fun:_ZN3gfx51_GLOBAL__N_gfx_codec_png_codec.cc_00000000_*
This is definitely bad.