Re: [cvs-nhc98] Re: [nhc-bugs] CVS copy doesn't compile
On Wed, Dec 19, 2001 at 10:54:22AM +0000, Malcolm Wallace wrote:
Fixed in CVS now.
Are you sure? I'm still getting it with a cleanly checked out version.
It is possible there is some lag between my checking in fixes and them appearing in the anonymous copy of the repository.
If you could find out then that might be useful.
And I needed to do
--- src/prelude/IO/IO.hs 2001/12/18 18:47:10 1.9 +++ src/prelude/IO/IO.hs 2001/12/18 22:13:48 @@ -78,7 +78,7 @@ import IoeGetHandle import IoeGetFileName
--- import HGetFileName +import HGetFileName
#if !defined(TRACING) import OpenSocket
in order to get that far this time.
That was one of the changes I already checked in, so it looks like
Just to be sure, did you comment out "import HGetFileName" or did you uncomment it?
there is indeed a delay on the repository. Try doing an update now.
"import HGetFileName" is still commented out and I asm still getting /home/ian/cvs/nhc/script/nhc98 -cpp -c +CTS -lib -redefine -CTS -P../PreludeIO -P../LowB/ -P../ix86-Linux -o /home/ian/cvs/nhc/targets/ix86-Linux/obj/prelude/IO/IO.o IO.hs ====== Errors when renaming: Identifier hGetFileName used at 28:6 is not defined. make[2]: *** [/home/ian/cvs/nhc/targets/ix86-Linux/obj/prelude/IO/IO.o] Error 1 make[2]: Leaving directory `/home/ian/cvs/nhc/src/prelude/IO' make[1]: *** [IO.make] Error 2 make[1]: Leaving directory `/home/ian/cvs/nhc/src/prelude' make: *** [targets/ix86-Linux/prelude] Error 2 Thanks Ian
It is possible there is some lag between my checking in fixes and them appearing in the anonymous copy of the repository.
If you could find out then that might be useful.
Hopefully the following little test will reveal whether that is the case.
Just to be sure, did you comment out "import HGetFileName" or did you uncomment it?
Aha. I didn't look closely enough at your diff. I uncommented the export of hGetFileName, but forgot to uncomment its corresponding import. For some reason, it compiled for me correctly anyway. So, try a cvs update in the src/prelude/IO directory now, and let's see if you get the newly commited file, or whether you need to wait. Regards, Malcolm
On Wed, Dec 19, 2001 at 11:54:08AM +0000, Malcolm Wallace wrote:
Just to be sure, did you comment out "import HGetFileName" or did you uncomment it?
Aha. I didn't look closely enough at your diff. I uncommented the export of hGetFileName, but forgot to uncomment its corresponding import. For some reason, it compiled for me correctly anyway.
So, try a cvs update in the src/prelude/IO directory now, and let's see if you get the newly commited file, or whether you need to wait.
I now see it uncommented :-) However, I am still hitting ----- /home/ian/cvs/nhc/script/nhc98 -cpp -p -c +CTS -part -redefine -CTS -P.. -P../PreludeIO -P../IO -prelude -o /home/ian/cvs/nhc/targets/ix86-Linux/objp/prelude/LowB/PrimOpenSocket.p.o PrimOpenSocket.hs Fail: GcodeFix.hs:165: Non-exhaustive patterns in function nthcon make[2]: *** [/home/ian/cvs/nhc/targets/ix86-Linux/objp/prelude/LowB/PrimOpenSocket.p.o] Error 1 make[2]: Leaving directory `/home/ian/cvs/nhc/src/prelude/LowB' make[1]: *** [LowB.make] Error 2 make[1]: Leaving directory `/home/ian/cvs/nhc/src/prelude' make: *** [targets/ix86-Linux/profprelude] Error 2 ----- or, with the debugging line uncommented (why is it commented out anyway?) Fail: nthcon: n==0 con==37 Thanks Ian
So, try a cvs update in the src/prelude/IO directory now, and let's see if you get the newly commited file, or whether you need to wait.
I now see it uncommented :-)
Good.
However, I am still hitting
Fail: GcodeFix.hs:165: Non-exhaustive patterns in function nthcon
Did you get the latest versions of include/IO.hi and include/IO.T.hi? The last 4 lines should read interface ! Prelude {-# NEED IO #-} newtype {-# #-} IO a; } instead of interface ! Prelude {-# NEED IO #-} data IO a; } This should fix it.
or, with the debugging line uncommented Fail: nthcon: n==0 con==37 (why is it commented out anyway?)
Don't know. It looks safe enough to keep it in. Regards, Malcolm
On Wed, Dec 19, 2001 at 12:22:55PM +0000, Malcolm Wallace wrote:
Did you get the latest versions of include/IO.hi and include/IO.T.hi?
Yes.
or, with the debugging line uncommented Fail: nthcon: n==0 con==37
con was 35 in the original problem. Thanks Ian
| > > or, with the debugging line uncommented | > > Fail: nthcon: n==0 con==37 | | con was 35 in the original problem. And I note that the error now occurs during the heap-profiling build, rather than the basic build. Investigating... Regards, Malcolm
participants (2)
-
Ian Lynagh -
Malcolm Wallace