Comments
Sort by recent activity
Hi Scott,
Does copying the PDB files make the framework methods appear in bold in the tree and grid? If so this means ANTS Profiler has recognised that source code information is available. You then just need to selected one of the methods and click on the link to tell it where to look for the source file (it should then find most of the others automatically). If it's not showing the methods in bold, then for whatever reason it's ignoring the PDB files. I'll have a chat with Andrew about it and see what the issue is likely to be.
Thanks,
Bart / comments
Hi Scott,
Does copying the PDB files make the framework methods appear in bold in the tree and grid? If so this means ANTS Profiler has recognised that source code information is available. You the...
Hi Dan,
Thanks for posting. The problem here is the lack of PDB files. Without PDB files there's no way to map back to any source code, where it exists. Now that we've acquired .NET Reflector we could potentially do something with the decompilation technology in this, however profiling the framework methods at this level of detail would lead to a significant performance penalty so you probably wouldn't want to do it all the time. So, certainly something we could consider adding in the future.
Thanks,
Bart / comments
Hi Dan,
Thanks for posting. The problem here is the lack of PDB files. Without PDB files there's no way to map back to any source code, where it exists. Now that we've acquired .NET Reflector we co...
Hi Scott,
If you have the PDBs you might try copying them alongside the framework assemblies as an interim fix. You might find this works (OTOH it might not if the PDBs don't contain line level offset mappings, but possibly worth a go). If it does you'll still need to point ANTS Profiler at the source files, although once you've done this for one file you should find it works for all of them.
Thanks,
Bart / comments
Hi Scott,
If you have the PDBs you might try copying them alongside the framework assemblies as an interim fix. You might find this works (OTOH it might not if the PDBs don't contain line level off...
Hi Paul,
This has now been fixed. Please refer to the early access build announcement in the main forum for download details: http://www.red-gate.com/MessageBoard/viewforum.php?f=86
Thanks,
Bart / comments
Hi Paul,
This has now been fixed. Please refer to the early access build announcement in the main forum for download details:http://www.red-gate.com/MessageBoard/viewforum.php?f=86
Thanks,
Bart
Hi Paul,
Whoops! Sorry about that. At least the red cross tells me where to go looking for the problem though. Thanks for posting.
Bart / comments
Hi Paul,
Whoops! Sorry about that. At least the red cross tells me where to go looking for the problem though. Thanks for posting.
Bart
Hi Paul,
I've tracked this down and fixed it, however as I said on another thread, various people are going on holiday at the moment, and we're moving version control systems, so it'll be a little while before I can make a patch build available. Please accept my apologies for this delay.
I'll let you know as soon as we have anything.
Thanks,
Bart / comments
Hi Paul,
I've tracked this down and fixed it, however as I said on another thread, various people are going on holiday at the moment, and we're moving version control systems, so it'll be a little ...
Hi Kirill,
Please refer to the early access build announcement in the main forum: http://www.red-gate.com/MessageBoard/viewforum.php?f=86
This should fix the problem you saw in the log file, although there are no fixes in the memory profiler. If you are still having issues with the memory profiler could you provide us with more details please.
Thanks,
Bart / comments
Hi Kirill,
Please refer to the early access build announcement in the main forum:http://www.red-gate.com/MessageBoard/viewforum.php?f=86
This should fix the problem you saw in the log file, althoug...
Sorry, did you try to load some memory profiler results with the performance profiler? / comments
Sorry, did you try to load some memory profiler results with the performance profiler?
Hi Kirill,
Thanks for reporting this. I've now (I hope) tracked down and fixed the error reported in the log file, however this is definitely a performance profiler log, so I'm still a little confused about whether this corresponds to the problem you had, or whether you had some other problem with the memory profiler (which AFAIK doesn't write a log file).
If you've any more information, or need anything else from us, please let me know. We'll put this fix in a 4.1 early access build, which we'll make available within the next few days. We'll announce it here and on my blog, so keep your eyes peeled.
Thanks,
Bart / comments
Hi Kirill,
Thanks for reporting this. I've now (I hope) tracked down and fixed the error reported in the log file, however this is definitely a performance profiler log, so I'm still a little confu...
Hi Paul,
Thanks for the offer of setting up a remote session; it might well prove useful. Let me have a chat to Stephen about it and we'll see if we can sort it out. Andrew, who wrote the click navigation, is away today so I'll ask him about this when he's back in the office.
The answer as to how we do it is a bit of both. Our source code view provides us with an AST for the currently loaded source file (for both C# and VB at any rate). From this we can grab all the methods, and for each method the execution info. We then try to stitch the execution info together with the code in the method based on method names, class names etc., but because we don't have a full code model for the profiled application this isn't 100% accurate. The problem is that without full type information for everything, any matching we do is necessarily going to be a bit on the fuzzy side.
It could be that there's just a problem resolving the target because there are two calls to a method with the same name, however that would be surprising because the parameter counts are different, and in any case the method calls would appear at different IL offsets, which would map to different lines in the source file. When you click on the version that takes a string, do you get a popup appearing with two different methods on it, or does it just take you straight to the source code?
Thanks,
Bart / comments
Hi Paul,
Thanks for the offer of setting up a remote session; it might well prove useful. Let me have a chat to Stephen about it and we'll see if we can sort it out. Andrew, who wrote the click nav...