Re: [Oorexx-devel] Experimenting security checks option
On Sun, Apr 26, 2009 at 5:05 PM, Jean-Louis Faucher wrote: > After reading this MSDN article > http://msdn.microsoft.com/en-us/library/aa289171.aspx > I decided to test these options, and modified orxwin32mak like that (added > security_check) : > The second security error ocurs when running SysDriveMap.testGroup. The > fifth and sixth call to SysDriveMap raises this error : "Run-Time Check > Failure #2 - Stack around the variable 'DeviceName' was corrupted" > In both cases, the call is self~assertNotNull(SysDriveMap(,'local')) > I can't find what's wrong with these two calls, and maybe I loose my time > for nothing, but... In rexxutil.cpp, the DeviceName variable is too small: CHAR DeviceName[3]; ... sprintf(DeviceName, "%c:\\", dnum+'A'-1); sprintf() will always append a null, so we have something like C:\ and no space for the null. sprintf() will place it at DeviceName[3] which is probably where the security error comes from, a buffer overrun. -- Mark Miesfeld -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
[Oorexx-devel] Experimenting security checks option
After reading this MSDN article http://msdn.microsoft.com/en-us/library/aa289171.aspx I decided to test these options, and modified orxwin32mak like that (added security_check) : !IF "$(NODEBUG)" == "1" security_check = /GS ... !ELSE security_check = /GS /RTCc /RTCs /RTCu ... !ENDIF cflags_common=/EHsc /nologo /D:_X86_ /DWIN32 $(VER_DEF) $(WARNING_FLAGS) -c $(my_cdebug) $(security_check) $(MK_ASM) $(RXDBG) /DNULL=0 I had my first security error when running the base\class\string testsuite, but it can be ignored : it's because of the /RTCc option which makes the runtime complain when casting to (char) in encodeBase64. I bypassed the errors by and-ing with 0xFF, as explained in the error message. The rest of the tests in base\class pass without error, including Stream.testGroup (I saw Ricks'call for test cases, I will try to write some, but probably not before next WE). concurency : ok directives : ok expressions : ok keyword : ok rexxutil : KO The second security error ocurs when running SysDriveMap.testGroup. The fifth and sixth call to SysDriveMap raises this error : "Run-Time Check Failure #2 - Stack around the variable 'DeviceName' was corrupted" In both cases, the call is self~assertNotNull(SysDriveMap(,'local')) I can't find what's wrong with these two calls, and maybe I loose my time for nothing, but... runtime.objects : ok source.file : ok special.variables : ok No security error when running Mike's getc.rex script... Jean-Louis -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] We desperately need some I/O test cases.
On Sun, Apr 26, 2009 at 3:27 PM, Mark Miesfeld wrote: > On Sun, Apr 26, 2009 at 3:15 PM, Rick McGuire wrote: > >> How about it folks, any volunteers >> out there willing to write some test cases for a problem area? > > Please do so folks. Oh, I forgot to add. Write the test cases and attach them as patches in Tracker. Or, you can even, under these special circumstances, just e-mail them straight to my gmail account. Which should be well known, I don't make much of a secret of it. -- Mark Miesfeld miesf...@gmail.com -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] We desperately need some I/O test cases.
On Sun, Apr 26, 2009 at 3:15 PM, Rick McGuire wrote: > I just fixed a crash in the stream class "query datetime" function on > windows that would never have gotten through if we had anything > resembling a regression suite for the I/O functions and class. We > really need at least some minimal tests for all of the various > functions, plus we need to have some tests for the different open > options (NOBUFFER, APPEND, etc.). Well, you're right about that. Guess I'd feel guilty if I wrote any ooDialog code today.I'll get started on some. > How about it folks, any volunteers > out there willing to write some test cases for a problem area? I Please do so folks. Don't worry about duplicating something I am doing. The test suite currently only takes a few minutes to run to completion with something like 16,000 tests. If 10 people out there each wrote 100 test cases that were similar to something I did, we'd just add them in and probably add less than 60 seconds to the total run time. Plus, similar is not the same. Most likely each test case would generate a slightly different code path. And that is what we need to catch things, a slightly different code path. -- Mark Miesfeld -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
[Oorexx-devel] We desperately need some I/O test cases.
I just fixed a crash in the stream class "query datetime" function on windows that would never have gotten through if we had anything resembling a regression suite for the I/O functions and class. We really need at least some minimal tests for all of the various functions, plus we need to have some tests for the different open options (NOBUFFER, APPEND, etc.). How about it folks, any volunteers out there willing to write some test cases for a problem area? I'm almost in favor of not shipping the final 4.0 version until we have a good regression suite, so the more hands we have helping with this, the more likely we might hit our end of May target for the beta. Rick -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
On Sat, Apr 25, 2009 at 8:23 PM, Mark Miesfeld wrote: > In working on the SysFileSearch bug, I tracked down what seems to be a > case where the VS 2005 compiler produces wrong code and the VS 2008 > produces code that works. Good thing I said "seems to be." I was 100% wrong on this, it was not the compiler at all but the programmers. ReadFile() takes a pointer to a 32-bit value to return actual count of bytes read. In the code clean up / move to 64-bit / conversion to the new API, the variable was changed to a size_t. In ReadFile() the variable was cast as a pointer to a 32-bit variable. Well, that's no good on 64-bit Windows. Anyhow, sorry I blamed the compiler, it wasn't a compiler bug at all. -- Mark Miesfeld -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] I/O again
> On Sun, Apr 26, 2009 at 11:01 AM, Rick McGuire wrote: > > > Mark, you might want to spin a new install package from the sandbox. > > I've fixed the query size/query datetime problems for directories. > > Actually I just started that, should be done shortly. Great ... send me the coords when done and I'll run with that from first thing tomorrow morning (UK time). Done for this evening, here ... fresh air/refreshment calls ... :-) Mike Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] I/O again
> I know exactly what you mean. The first few times I saw the error > > 'RAY' is not recognized as an internal or external command, > operable program or batch file. > > I thought I'd just mistyped something. I don't even have any theories > on how/why this is happenng based on the symptoms we're seeing. I > have enough debug evidence from the times I have been able to > reproduce this to be very certain the bad character is coming directly > from the CRT _read() function. At least in my situation, everything > was single threaded. Yes, this is pretty much the 'worst nightmare' bug .. probably someone else's fault, cannot be easily reproduced, but all too obvious when it happens (and can have weird side-effects, when the change to an 'r' meant an old program I hadn't run in 10+ years and had completely forgotten about got run and sent off an e-mail to an old acquaintance ...). > I do have a questions for you Mike. I've only ever seen this on a > first read. I have most often (60% of times?) seen it there (as today) but have also seen in happen much later, and at least once on two consecutive reads in the same elap session. >Have you ever seen the overlay occur in any place other > than the first character or a line? No. Always only the first character, and always 'r'. > Have you noticed any pattern as > to things you might have done *before* the command with the overlay? No. Most often it is the first command in a new command window, but it also happens later. I thought for a while that it happened more often with commands starting with 'c' but no longer think that is the case. I also thought that perhaps it needed two command windows open .. but again, the case today was definitely the only command window open (because if I have one left open during reinstall the install process alerts), and it was the first Rexx program run after re-installing ooRexx. Mike [so much for my new resolution to not check IBM mail on a Sunday! :-)))] Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] I/O again
On Sun, Apr 26, 2009 at 11:01 AM, Rick McGuire wrote: > Mark, you might want to spin a new install package from the sandbox. > I've fixed the query size/query datetime problems for directories. Actually I just started that, should be done shortly. -- Mark Miesfeld -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] I/O again
I know exactly what you mean. The first few times I saw the error 'RAY' is not recognized as an internal or external command, operable program or batch file. I thought I'd just mistyped something. I don't even have any theories on how/why this is happenng based on the symptoms we're seeing. I have enough debug evidence from the times I have been able to reproduce this to be very certain the bad character is coming directly from the CRT _read() function. At least in my situation, everything was single threaded. I do have a questions for you Mike. I've only ever seen this on a first read. Have you ever seen the overlay occur in any place other than the first character or a line? Have you noticed any pattern as to things you might have done *before* the command with the overlay? Mark, you might want to spin a new install package from the sandbox. I've fixed the query size/query datetime problems for directories. Rick On Sun, Apr 26, 2009 at 12:52 PM, Mike Cowlishaw wrote: > > Yeah .. sorry about that. I was really hoping VS2008 fixed it, but always > knew that 'not seeing the dreaded r' was a negative report, not a definitive > one. Darn thing seems almost determined to hide itself. I'm just glad you > (Rick) saw it too, as I'm sure no one would believe me otherwise! > > Mike > > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - > Mike Cowlishaw, IBM Fellow > http://bit.ly/mfc > IBM UK (MP8), PO Box 31, Birmingham Road, Warwick, CV34 5JL > > Rick McGuire wrote on 26/04/2009 13:31:02: > >> Rick McGuire >> 26/04/2009 13:31 >> >> Please respond to >> Open Object Rexx Developer Mailing List >> >> >> To >> >> Open Object Rexx Developer Mailing List >> >> >> cc >> >> Subject >> >> Re: [Oorexx-devel] I/O again >> >> Sigh, this is why I was really hoping the 2008 conversion would make >> the problem go away. I figured it would take a couple of iterations >> to get all of the conversion bugs worked out. I think I know how to >> fix this one, so I'll probably have a fix available shortly. >> >> Rick >> >> On Sun, Apr 26, 2009 at 8:15 AM, Mike Cowlishaw wrote: >> > >> > Well, installed latest (Rick new file I/O) but unfortunately had to back >> > it >> > out almost immediately, because now stream 'query datetime' and 'query >> > size' >> > both return empty string for a directory. The former I can work around >> > (and >> > did when it changed a week or two ago), but the latter I have several >> > hundred instances of >> > >> > So I then went back to Mark's VS 2008 version ... >> > >> > ... and immediate got a 'r' overlay, on the first parse pull after >> > install. >> > >> > >> > Staying with that one for now, however. >> > >> > Mike >> > >> > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - >> > Mike Cowlishaw, IBM Fellow >> > http://bit.ly/mfc >> > IBM UK (MP8), PO Box 31, Birmingham Road, Warwick, CV34 5JL >> > >> > >> > >> > >> > Unless stated otherwise above: >> > IBM United Kingdom Limited - Registered in England and Wales with number >> > 741598. >> > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 >> > 3AU >> > >> > >> > >> > >> > >> > >> >> -- >> > Crystal Reports - New Free Runtime and 30 Day Trial >> > Check out the new simplified licensign option that enables unlimited >> > royalty-free distribution of the report engine for externally facing >> > server and web deployment. >> > http://p.sf.net/sfu/businessobjects >> > ___ >> > Oorexx-devel mailing list >> > Oorexx-devel@lists.sourceforge.net >> > https://lists.sourceforge.net/lists/listinfo/oorexx-devel >> > >> > >> >> >> -- >> Crystal Reports - New Free Runtime and 30 Day Trial >> Check out the new simplified licensign option that enables unlimited >> royalty-free distribution of the report engine for externally facing >> server and web deployment. >> http://p.sf.net/sfu/businessobjects >> ___ >> Oorexx-devel mailing list >> Oorexx-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/oorexx-devel > > > > > > Unless stated otherwise above: > IBM United Kingdom Limited - Registered in England and Wales with number > 741598. > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU > > > > > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-deve
Re: [Oorexx-devel] I/O again
Yeah .. sorry about that. I was really hoping VS2008 fixed it, but always knew that 'not seeing the dreaded r' was a negative report, not a definitive one. Darn thing seems almost determined to hide itself. I'm just glad you (Rick) saw it too, as I'm sure no one would believe me otherwise! Mike - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Mike Cowlishaw, IBM Fellow http://bit.ly/mfc IBM UK (MP8), PO Box 31, Birmingham Road, Warwick, CV34 5JL Rick McGuire wrote on 26/04/2009 13:31:02: > Rick McGuire > 26/04/2009 13:31 > > Please respond to > Open Object Rexx Developer Mailing List > > To > > Open Object Rexx Developer Mailing List > > cc > > Subject > > Re: [Oorexx-devel] I/O again > > Sigh, this is why I was really hoping the 2008 conversion would make > the problem go away. I figured it would take a couple of iterations > to get all of the conversion bugs worked out. I think I know how to > fix this one, so I'll probably have a fix available shortly. > > Rick > > On Sun, Apr 26, 2009 at 8:15 AM, Mike Cowlishaw wrote: > > > > Well, installed latest (Rick new file I/O) but unfortunately had to back it > > out almost immediately, because now stream 'query datetime' and 'query size' > > both return empty string for a directory. The former I can work around (and > > did when it changed a week or two ago), but the latter I have several > > hundred instances of > > > > So I then went back to Mark's VS 2008 version ... > > > > ... and immediate got a 'r' overlay, on the first parse pull after install. > > > > > > Staying with that one for now, however. > > > > Mike > > > > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - > > Mike Cowlishaw, IBM Fellow > > http://bit.ly/mfc > > IBM UK (MP8), PO Box 31, Birmingham Road, Warwick, CV34 5JL > > > > > > > > > > Unless stated otherwise above: > > IBM United Kingdom Limited - Registered in England and Wales with number > > 741598. > > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU > > > > > > > > > > > > > -- > > Crystal Reports - New Free Runtime and 30 Day Trial > > Check out the new simplified licensign option that enables unlimited > > royalty-free distribution of the report engine for externally facing > > server and web deployment. > > http://p.sf.net/sfu/businessobjects > > ___ > > Oorexx-devel mailing list > > Oorexx-devel@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/oorexx-devel > > > > > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/oorexx-devel Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
On Sun, Apr 26, 2009 at 6:58 AM, Rick McGuire wrote: > Did you forget to paste the link? http://msdn.microsoft.com/en-us/library/ms246609.aspx Didn't have my coffee yet. -- Mark Miesfeld -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
Did you forget to paste the link? On Sun, Apr 26, 2009 at 9:45 AM, Mark Miesfeld wrote: > On Sun, Apr 26, 2009 at 4:49 AM, Rick McGuire wrote: > >> It's starting to look like upgrading to VS 2008 might be a winner for >> this. I'm more willing to make this change than switch to the new >> version of SysFile. Can the two version coexist on the same machine, >> or do I need to uninstall 2005 first? > > Here's a link that says you can and has some info on things to do / > watch out for. > > I did'nt think you could, but I guess you can since the source is Microsoft. > > -- > Mark Miesfeld > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/oorexx-devel > -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
On Sun, Apr 26, 2009 at 4:49 AM, Rick McGuire wrote: > It's starting to look like upgrading to VS 2008 might be a winner for > this. I'm more willing to make this change than switch to the new > version of SysFile. Can the two version coexist on the same machine, > or do I need to uninstall 2005 first? Here's a link that says you can and has some info on things to do / watch out for. I did'nt think you could, but I guess you can since the source is Microsoft. -- Mark Miesfeld -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
On Sun, Apr 26, 2009 at 4:49 AM, Rick McGuire wrote: > It's starting to look like upgrading to VS 2008 might be a winner for > this. I'm more willing to make this change than switch to the new > version of SysFile. I think upgrading to VS 2008 has less potential for letting a bug slip through to the release than switching to SysFile. I've been working with builds from both VS 2008 and VS 2005, it would just depend on which system I was on at the time -- Mark Miesfeld -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] I/O again
Sigh, this is why I was really hoping the 2008 conversion would make the problem go away. I figured it would take a couple of iterations to get all of the conversion bugs worked out. I think I know how to fix this one, so I'll probably have a fix available shortly. Rick On Sun, Apr 26, 2009 at 8:15 AM, Mike Cowlishaw wrote: > > Well, installed latest (Rick new file I/O) but unfortunately had to back it > out almost immediately, because now stream 'query datetime' and 'query size' > both return empty string for a directory. The former I can work around (and > did when it changed a week or two ago), but the latter I have several > hundred instances of > > So I then went back to Mark's VS 2008 version ... > > ... and immediate got a 'r' overlay, on the first parse pull after install. > > > Staying with that one for now, however. > > Mike > > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - > Mike Cowlishaw, IBM Fellow > http://bit.ly/mfc > IBM UK (MP8), PO Box 31, Birmingham Road, Warwick, CV34 5JL > > > > > Unless stated otherwise above: > IBM United Kingdom Limited - Registered in England and Wales with number > 741598. > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU > > > > > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/oorexx-devel > > -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] I/O again
Well, installed latest (Rick new file I/O) but unfortunately had to back it out almost immediately, because now stream 'query datetime' and 'query size' both return empty string for a directory. The former I can work around (and did when it changed a week or two ago), but the latter I have several hundred instances of So I then went back to Mark's VS 2008 version ... ... and immediate got a 'r' overlay, on the first parse pull after install. Staying with that one for now, however. Mike - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Mike Cowlishaw, IBM Fellow http://bit.ly/mfc IBM UK (MP8), PO Box 31, Birmingham Road, Warwick, CV34 5JL Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Questions ad an object being used by a different RexxInstance ...
Rick McGuire wrote: > I'm not sure this question makes any sense. An interaction on an > object occurrs in the context of where the interaction is triggered. > Objects are just data, and have no knowledge of where they have been > created. If one instance stores an object in .environment (generally > a bad practice), then it is just an object that's available for any > instance. It will remain alive until the global Rexx environment is > terminated. > The reason for asking is this: if having multiple RexxInterpreterInstances running for different Java objects, and if one wishes to allow Java to interact with proxy Rexx objects (in the JNI layer), do I have to use the cached RexxInterpreterInstance (in the JNI layer) from which the proxied Rexx objects come from originally, or is it o.k. to use any present RexxInterpreterInstance to attach to and then forwarding the Rexx proxy object messages from Java. This is about contemplating of allowing Java to interact with Rexx objects in a general way (not only for event handlers) and sending them messages. If I understood your answer correctly, then it does not matter which RexxInterpreterInstance I would use to send messages to cached (in the JNI layer) Rexx objects. (Also assuming, that once all RexxInterpreterInstances got terminated, no such invocation should be allowed anymore, but throw an exception.) ---rony -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
It's starting to look like upgrading to VS 2008 might be a winner for this. I'm more willing to make this change than switch to the new version of SysFile. Can the two version coexist on the same machine, or do I need to uninstall 2005 first? Rick On Sat, Apr 25, 2009 at 11:23 PM, Mark Miesfeld wrote: > Rick, > > In working on the SysFileSearch bug, I tracked down what seems to be a > case where the VS 2005 compiler produces wrong code and the VS 2008 > produces code that works. > > It's bizarre, so it takes a bit to explain it. On Vista 64-bit, the > GetLine() function causes a memory exception here: > > scan = (char *)memchr(filedata->scan, CH_NL, filedata->data); > > The exception comes because it has gone past the end of the file > buffer. The check for the end of the buffer is right before, and the > whole thing looks like: > > if (!(filedata->data)) { /* if out of current buffer */ > if (filedata->residual) { /* may be another buffer */ > ReadNextBuffer(filedata); /* try to read one */ > if (!filedata->data) /* nothing more? */ > return true; /* all done */ > } > else > return true; /* return EOF condition */ > } > /* look for a carriage return */ > scan = (char *)memchr(filedata->scan, CH_NL, filedata->data); > > filedata->data keeps track of where in the buffer we are and when it > is 0, we've used up the buffer. > > The bizarre thing is, when I compile with Visual Studio 2005, when > filedata->data == 0, the if !(filedata->data) test passes and the > function keeps going on past the end of the buffer. Eventually, > memchr goes to compare memory it can't access. > > If I compile with Visual Studio 2008, then things work as they should. > This is with the no debug build only. The debug build works fine > when compiled with either VS 2005 or VS 2008. > > I sent you a text file with a lot output. It is easier to read when > the lines don't wrap. > > I used this printf to show this behaviour: > > printf("GetLine: size=%u scan=%p filedata->data=%u residual=%u > filedata->data == 0 ? %d\n", > filedata->size, filedata->scan, filedata->data, > filedata->residual, (filedata->data == 0)); > > > For the build that crashes you see this: > > GetLine: size=31452 scan=001E9FDC filedata->data=0 residual=0 > filedata->data == 0 ? 0 > GetLine: size=31452 scan=001EE48D filedata->data=4294949711 > residual=0 filedata->data == 0 ? 0 > > You see that even though filedata->data does equal 0, (filedata->data > == 0) returns 0. > > So, SysFileSearch keeps calling GetLine() until you finally reach > memory that is not accessible. > > This only happens on Vista 64-bit, only on a no debug build, and only > when the compiler is VS 2005. On XP 64-bit it works fine with either > compiler. Also on 32-bit Vista it works fine with either compiler. > > If there was the same problem on Vista 32 bit, i.e. it crashed with > the VS 2005 compiled version but worked with the VS 2008, then I was > thinking maybe a similar thing is happening with Mike's problem. > Still, if Mike can not reproduce his problem using the VS 2008 > compiled version, then maybe it is a similar thing. > > -- > Mark Miesfeld > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/oorexx-devel > -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Questions ad an object being used by a different RexxInstance ...
I'm not sure this question makes any sense. An interaction on an object occurrs in the context of where the interaction is triggered. Objects are just data, and have no knowledge of where they have been created. If one instance stores an object in .environment (generally a bad practice), then it is just an object that's available for any instance. It will remain alive until the global Rexx environment is terminated. Rick On Sun, Apr 26, 2009 at 6:32 AM, Rony G. Flatscher wrote: > Given two different RexxInstances R1 and R2 running in the same process. In > R1 an object O1 is created and saved in ".environment". > > Two questions ad RexxInterpreter context: > > If from R2 an interaction occurs with O1 (addressed via ".environment"), in > which context does O1 execute, in R1 (the creator RexxInterpreterInstance) > or in R2 (the invoker RexxInterpreterInstance)? > If R1 is terminated, but R2 interacts with O1 (addressed via > ".environment"), in which context does O1 execute, in R1 (the creating > RexxInterpreterInstance, but terminated) or in R2 (the invoking > RexxInterpreterInstance)? > > ---rony > > > > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/oorexx-devel > > -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
This looks like a classic optimizer bug to me. For some reason, the code generator has obviously decided it can used a cached version of the data pointer for the if test rather than reloadingbut it appears to make that decision only for the if test. It might be interesting to try declaring that value as volatile and see if things change. Rick On Sat, Apr 25, 2009 at 11:23 PM, Mark Miesfeld wrote: > Rick, > > In working on the SysFileSearch bug, I tracked down what seems to be a > case where the VS 2005 compiler produces wrong code and the VS 2008 > produces code that works. > > It's bizarre, so it takes a bit to explain it. On Vista 64-bit, the > GetLine() function causes a memory exception here: > > scan = (char *)memchr(filedata->scan, CH_NL, filedata->data); > > The exception comes because it has gone past the end of the file > buffer. The check for the end of the buffer is right before, and the > whole thing looks like: > > if (!(filedata->data)) { /* if out of current buffer */ > if (filedata->residual) { /* may be another buffer */ > ReadNextBuffer(filedata); /* try to read one */ > if (!filedata->data) /* nothing more? */ > return true; /* all done */ > } > else > return true; /* return EOF condition */ > } > /* look for a carriage return */ > scan = (char *)memchr(filedata->scan, CH_NL, filedata->data); > > filedata->data keeps track of where in the buffer we are and when it > is 0, we've used up the buffer. > > The bizarre thing is, when I compile with Visual Studio 2005, when > filedata->data == 0, the if !(filedata->data) test passes and the > function keeps going on past the end of the buffer. Eventually, > memchr goes to compare memory it can't access. > > If I compile with Visual Studio 2008, then things work as they should. > This is with the no debug build only. The debug build works fine > when compiled with either VS 2005 or VS 2008. > > I sent you a text file with a lot output. It is easier to read when > the lines don't wrap. > > I used this printf to show this behaviour: > > printf("GetLine: size=%u scan=%p filedata->data=%u residual=%u > filedata->data == 0 ? %d\n", > filedata->size, filedata->scan, filedata->data, > filedata->residual, (filedata->data == 0)); > > > For the build that crashes you see this: > > GetLine: size=31452 scan=001E9FDC filedata->data=0 residual=0 > filedata->data == 0 ? 0 > GetLine: size=31452 scan=001EE48D filedata->data=4294949711 > residual=0 filedata->data == 0 ? 0 > > You see that even though filedata->data does equal 0, (filedata->data > == 0) returns 0. > > So, SysFileSearch keeps calling GetLine() until you finally reach > memory that is not accessible. > > This only happens on Vista 64-bit, only on a no debug build, and only > when the compiler is VS 2005. On XP 64-bit it works fine with either > compiler. Also on 32-bit Vista it works fine with either compiler. > > If there was the same problem on Vista 32 bit, i.e. it crashed with > the VS 2005 compiled version but worked with the VS 2008, then I was > thinking maybe a similar thing is happening with Mike's problem. > Still, if Mike can not reproduce his problem using the VS 2008 > compiled version, then maybe it is a similar thing. > > -- > Mark Miesfeld > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/oorexx-devel > -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects ___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
[Oorexx-devel] Questions ad an object being used by a different RexxInstance ...
Given two different RexxInstances R1 and R2 running in the same process. In R1 an object O1 is created and saved in ".environment". Two questions ad RexxInterpreter context: 1. If from R2 an interaction occurs with O1 (addressed via ".environment"), in which context does O1 execute, in R1 (the creator RexxInterpreterInstance) or in R2 (the invoker RexxInterpreterInstance)? 2. If R1 is terminated, but R2 interacts with O1 (addressed via ".environment"), in which context does O1 execute, in R1 (the creating RexxInterpreterInstance, but terminated) or in R2 (the invoking RexxInterpreterInstance)? ---rony -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects___ Oorexx-devel mailing list Oorexx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oorexx-devel
Re: [Oorexx-devel] Tracked down problem with: compiler, Vista, ?
Interesting ... would be interesting to see the asm generated in the good/bad cases (/FAs option, I think)... Only 64-bit Vista is a bit strange though (I'm on Vista 32, and Rick also saw the problem on XP). However, I think you may well be right about something bad in the 2005 compiler or libraries -- ran with the 2008 version for Friday evening, all of yesterday, and a hour or so this morning -- no pesky 'r' seen. Will switch to Rick's new version later this morning. Mike - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Mike Cowlishaw, IBM Fellow http://bit.ly/mfc IBM UK (MP8), PO Box 31, Birmingham Road, Warwick, CV34 5JL Mark Miesfeld wrote on 26/04/2009 04:23:04: > Mark Miesfeld > 26/04/2009 04:23 > > Please respond to > Open Object Rexx Developer Mailing List > > To > > ooRexx Dev List > > cc > > Subject > > [Oorexx-devel] Tracked down problem with: compiler, Vista, ? > > Rick, > > In working on the SysFileSearch bug, I tracked down what seems to be a > case where the VS 2005 compiler produces wrong code and the VS 2008 > produces code that works. > > It's bizarre, so it takes a bit to explain it. On Vista 64-bit, the > GetLine() function causes a memory exception here: > > scan = (char *)memchr(filedata->scan, CH_NL, filedata->data); > > The exception comes because it has gone past the end of the file > buffer. The check for the end of the buffer is right before, and the > whole thing looks like: > > if (!(filedata->data)) { /* if out of current buffer */ > if (filedata->residual) { /* may be another buffer */ > ReadNextBuffer(filedata);/* try to read one*/ > if (!filedata->data) /* nothing more? */ > return true; /* all done */ > } > else > return true; /* return EOF condition */ > } >/* look for a carriage return */ > scan = (char *)memchr(filedata->scan, CH_NL, filedata->data); > > filedata->data keeps track of where in the buffer we are and when it > is 0, we've used up the buffer. > > The bizarre thing is, when I compile with Visual Studio 2005, when > filedata->data == 0, the if !(filedata->data) test passes and the > function keeps going on past the end of the buffer. Eventually, > memchr goes to compare memory it can't access. > > If I compile with Visual Studio 2008, then things work as they should. > This is with the no debug build only. The debug build works fine > when compiled with either VS 2005 or VS 2008. > > I sent you a text file with a lot output. It is easier to read when > the lines don't wrap. > > I used this printf to show this behaviour: > > printf("GetLine: size=%u scan=%p filedata->data=%u residual=%u > filedata->data == 0 ? %d\n", >filedata->size, filedata->scan, filedata->data, > filedata->residual, (filedata->data == 0)); > > > For the build that crashes you see this: > > GetLine: size=31452 scan=001E9FDC filedata->data=0 residual=0 > filedata->data == 0 ? 0 > GetLine: size=31452 scan=001EE48D filedata->data=4294949711 > residual=0 filedata->data == 0 ? 0 > > You see that even though filedata->data does equal 0, (filedata->data > == 0) returns 0. > > So, SysFileSearch keeps calling GetLine() until you finally reach > memory that is not accessible. > > This only happens on Vista 64-bit, only on a no debug build, and only > when the compiler is VS 2005. On XP 64-bit it works fine with either > compiler. Also on 32-bit Vista it works fine with either compiler. > > If there was the same problem on Vista 32 bit, i.e. it crashed with > the VS 2005 compiled version but worked with the VS 2008, then I was > thinking maybe a similar thing is happening with Mike's problem. > Still, if Mike can not reproduce his problem using the VS 2008 > compiled version, then maybe it is a similar thing. > > -- > Mark Miesfeld > > -- > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensign option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ___ > Oorexx-devel mailing list > Oorexx-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/oorexx-devel Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU -- Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensign option that enables unlimited royalty-free distribution of the report engine for externally facing se