- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello,
Since IF7.1 was first released a month ago, my program, which compiles perfectly well with CVF6.6 and IF7.0, has be giving me "Internal Compiler Error-Please notify Intel"
I have done so 3 times in the last month and supplied test code with which Intel Premier Support was able to reproduce the error. I was told that the error had been identified and would be fixed in this (the third) release. I still get the error msg. Since, it has not been fixed, I conclude that it must be something quirky with my code, or else many people would be complaining. I have rearranged the order of compilation of Modules, thinking that it couldn't handle a file in which the first module "USED" the second and a few other legal, but not too common constructs (e.g. a function in a module returning an array). Has anyone else had similar problems or can offer a suggestion as to where to look? Intel support has been polite, responsive and completely useless. Since my motivation for buying IF7 was the likely loss of technical support and bug fixes for CVF6.6 I can't help but wonder if I should just stick with it until IA32 processors are obsolete. (Sorry for venting my frustration.)
Thanks and regards,
Keith Keefer
Since IF7.1 was first released a month ago, my program, which compiles perfectly well with CVF6.6 and IF7.0, has be giving me "Internal Compiler Error-Please notify Intel"
I have done so 3 times in the last month and supplied test code with which Intel Premier Support was able to reproduce the error. I was told that the error had been identified and would be fixed in this (the third) release. I still get the error msg. Since, it has not been fixed, I conclude that it must be something quirky with my code, or else many people would be complaining. I have rearranged the order of compilation of Modules, thinking that it couldn't handle a file in which the first module "USED" the second and a few other legal, but not too common constructs (e.g. a function in a module returning an array). Has anyone else had similar problems or can offer a suggestion as to where to look? Intel support has been polite, responsive and completely useless. Since my motivation for buying IF7 was the likely loss of technical support and bug fixes for CVF6.6 I can't help but wonder if I should just stick with it until IA32 processors are obsolete. (Sorry for venting my frustration.)
Thanks and regards,
Keith Keefer
Link Copied
3 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
CVF support is still available.
If you'll give me the report numbers of the problems you've submitted. I'll take another look.
Steve
If you'll give me the report numbers of the problems you've submitted. I'll take another look.
Steve
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Steve,
Thanks. I waited until I thought the problem might have some more general interest before posting to the Forum, not wanting to use it in lieu of Intel support. The Issue number is 183310, under my name. The original source files were included. I've tried rewriting the code to "divide and conquer" to no avail. I have an Include line for MKL in my ifl.cfg and turn off warnings
because I use tab characters (you might suggest that Intel put in a switch to ignore tabs-we former DEC programmers still tend to use them-an anachronism from PDP-11 Fortran).If the old sources cause problems, I send some new versions, which also give the compiler error.
I've never needed tech support for C(D)VF :).
Thanks and regards
Keith
Thanks. I waited until I thought the problem might have some more general interest before posting to the Forum, not wanting to use it in lieu of Intel support. The Issue number is 183310, under my name. The original source files were included. I've tried rewriting the code to "divide and conquer" to no avail. I have an Include line for MKL in my ifl.cfg and turn off warnings
because I use tab characters (you might suggest that Intel put in a switch to ignore tabs-we former DEC programmers still tend to use them-an anachronism from PDP-11 Fortran).If the old sources cause problems, I send some new versions, which also give the compiler error.
I've never needed tech support for C(D)VF :).
Thanks and regards
Keith
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Keith,
I think I see the confusion here.
Feilong, the Intel support tech, told you that the problem was fixed in compilers released after 4/25/03 and "the next release should fix your problem." You apparently misunderstood this as saying that if you downloaded what was out there on premier.intel.com now that you'd get the fix. This is not the case. As far as I can tell, a compiler with this fix has not yet been released.
The easy way to tell what "build date" you have (which is what Feilong really should have said rather than "released after") is to type the ifl command from a command prompt with no arguments. The compiler will identify itself with a line that ends with text similar to this: "Build 20030410Z". In this example (which is the latest one I have on my system), the build date is 4/10/03 and it does not contain the fix for your problem (which I verified, I get the error too.)
I looked at Premier Support and see that there is a 7.1.010 version that was released on 5/1, but if you look at the build date in the description, it says 20030424, meaning that your fix just missed this release. You'll have to wait for the next one, which should be within a couple of weeks. Sorry for the confusion and inconvenience.
Steve
I think I see the confusion here.
Feilong, the Intel support tech, told you that the problem was fixed in compilers released after 4/25/03 and "the next release should fix your problem." You apparently misunderstood this as saying that if you downloaded what was out there on premier.intel.com now that you'd get the fix. This is not the case. As far as I can tell, a compiler with this fix has not yet been released.
The easy way to tell what "build date" you have (which is what Feilong really should have said rather than "released after") is to type the ifl command from a command prompt with no arguments. The compiler will identify itself with a line that ends with text similar to this: "Build 20030410Z". In this example (which is the latest one I have on my system), the build date is 4/10/03 and it does not contain the fix for your problem (which I verified, I get the error too.)
I looked at Premier Support and see that there is a 7.1.010 version that was released on 5/1, but if you look at the build date in the description, it says 20030424, meaning that your fix just missed this release. You'll have to wait for the next one, which should be within a couple of weeks. Sorry for the confusion and inconvenience.
Steve

Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page