Re: Resend pdh[3:3]: remove unnecessary test done in PdhLookupPerfNameByIndexW

[Hans Leidekker]

On Thursday 14 February 2008 09:57:32 am Jeff Latimer wrote:
> +    machineW = pdh_strdup_aw( machine );
> +
> +    if (!(ret = PdhLookupPerfNameByIndexW( machineW, index, bufferW, &sizeW )))

In the unlikely event that pdh_strdup_aw fails the query will be run on
the local machine instead of the remote machine. Please check for allocation
failures like most other functions in there do.

 -Hans

Resend pdh[3:3]: remove unnecessary test done in PdhLookupPerfNameByIndexW

[Jeff Latimer]

Hans Leidekker wrote:
> On Thursday 14 February 2008 09:57:32 am Jeff Latimer wrote:
> > +    machineW = pdh_strdup_aw( machine );
> > +
> > +    if (!(ret = PdhLookupPerfNameByIndexW( machineW, index, bufferW, &sizeW )))
>
> In the unlikely event that pdh_strdup_aw fails the query will be run on
> the local machine instead of the remote machine. Please check for allocation
> failures like most other functions in there do.
>
> -Hans

I looked at that and it required another test.  When I discussed this 
last with Dmitry, I said it would be best to pass machine and he agreed 
that this would be best.  The checking for alloc failures will generate 
similar amounts of code to what I am removing.  I am happy to do it but 
have a concern that the code will look to be inefficient and not 
accepted.  Logically, the patch is the correct way to proceed as machine 
should be passed even if not implemented yet.  That said, if the patch 
is acceptable, there is another case that will need to be brought into line.

Jeff

Resend pdh[3:3]: remove unnecessary test done inPdhLookupPerfNameByIndexW

[Dmitry Timoshkov]

"Jeff Latimer" <lats(a)yless4u.com.au> wrote:
> I looked at that and it required another test.  When I discussed this 
> last with Dmitry, I said it would be best to pass machine and he agreed 
> that this would be best.  The checking for alloc failures will generate 
> similar amounts of code to what I am removing.  I am happy to do it but 
> have a concern that the code will look to be inefficient and not 
> accepted.  Logically, the patch is the correct way to proceed as machine 
> should be passed even if not implemented yet.  That said, if the patch 
> is acceptable, there is another case that will need to be brought into line.

Hans means something like the following done in PdhOpenQueryA:

WCHAR *sourceW = NULL;
if (source && !(sourceW = pdh_strdup_aw( source ))) return PDH_MEMORY_ALLOCATION_FAILURE;

and that's certainly not a large amount of code. Actually all pdh_strdup_aw
callers already check for a possible NULL, but that may be sent as a separate
patch IMO.

-- 
Dmitry.