- Cvs-hugs - Haskell.org

cvs commit: hugs98/lib/exts .cvsignore
by Alastair Reid 20 Jun '02

20 Jun '02

reid 2002/06/20 09:15:36 PDT Added files: lib/exts .cvsignore Log: Forget the ffi

1 0

cvs commit: hugs98/tests/ffi .cvsignore
by Alastair Reid 20 Jun '02

20 Jun '02

reid 2002/06/20 09:10:31 PDT Modified files: tests/ffi .cvsignore Log: More thoroughly ignore intermediate files Revision Changes Path 1.2 +3 -1 hugs98/tests/ffi/.cvsignore

1 0

RE: cvs commit: hugs98/src HsFFI.h builtin.c connect.h ffi.c static.c
by Simon Marlow 19 Jun '02

19 Jun '02

> reid 2002/06/17 14:46:48 PDT > > Modified files: > src HsFFI.h builtin.c connect.h ffi.c > static.c > Log: > Made dynamic and wrapper forms work. > > This is my current test program: Thanks Alastair, I slurped your test program into GHC's test suite. I had to make a few changes to get it to run: the main one is that you used 'Ptr Char' in a lot of places where the type should have been 'Ptr CChar', i.e. CString. I guess in Hugs that doesn't matter too much because Char and CChar have the same representation at the C level, but that's not the case for GHC. I've attached my version. You might want to take a look in fptools/testsuite/tests/ghc-regress/ccall, and see if there's anything in there that might be useful for your test suite. Cheers, Simon

1 0

cvs commit: hugs98/src/unix aclocal.m4 configure configure.in
by Alastair Reid 18 Jun '02

18 Jun '02

reid 2002/06/18 15:40:19 PDT Modified files: src/unix aclocal.m4 configure configure.in Log: Stole some autoconfery from fptools/{aclocal.m4,configure.in}. Thanks Dudes! Revision Changes Path 1.6 +191 -0 hugs98/src/unix/aclocal.m4 1.23 +4205 -159 hugs98/src/unix/configure 1.26 +85 -0 hugs98/src/unix/configure.in

1 0

cvs commit: hugs98/lib/exts Storable.hs
by Alastair Reid 18 Jun '02

18 Jun '02

reid 2002/06/18 13:28:25 PDT Modified files: lib/exts Storable.hs Log: Replaced libspec "[Storable_aux.o]" with "[Storable_aux.c]" Revision Changes Path 1.4 +51 -51 hugs98/lib/exts/Storable.hs

1 0

cvs commit: hugs98/tests testScript hugs98/tests/ffi FileIO.hs FileIO.input FileIO.output String.hs String.input String.output
by Alastair Reid 18 Jun '02

18 Jun '02

reid 2002/06/18 13:26:55 PDT Modified files: tests testScript Added files: tests/ffi FileIO.hs FileIO.input FileIO.output String.hs String.input String.output Log: More runtime checks. I'm not sure how portable these are - I expect we'll iterate a few times to find something portable. You should expect to see some spurious warnings from gcc as it complains about perfectly legal but unusual code (like giving a function prototype for gcc builtin functions). The output I see from these tests is: ./../src/hugs +q +w -pHugs: -P../lib:../lib/hugs:../lib/exts +G ffi/Sin.hs < /dev/null expected stderr not matched by reality *** /tmp/no_stderr27731 Tue Jun 18 21:34:24 2002 --- /tmp/runtest27731.2 Tue Jun 18 21:34:24 2002 *************** *** 0 **** --- 1,2 ---- + Sin.c: In function `hugsprim_sin_3': + Sin.c:51: warning: built-in function `sin' declared as non-function -- !!! Testing static, dynamic and wrapped import of trig function -- !!! Testing marshalling of strings ./../src/hugs +q +w -pHugs: -P../lib:../lib/hugs:../lib/exts +G ffi/String.hs < /dev/null expected stderr not matched by reality *** /tmp/no_stderr27767 Tue Jun 18 21:34:25 2002 --- /tmp/runtest27767.2 Tue Jun 18 21:34:25 2002 *************** *** 0 **** --- 1,3 ---- + MarshalUtils.c: In function `hugsprim_memcpy_0': + MarshalUtils.c:22: warning: conflicting types for built-in function `memcpy' + MarshalUtils.c:22: warning: extern declaration of `memcpy' doesn't match global one -- !!! Testing marshalling of strings -- !!! Testing system calls ./../src/hugs +q +w -pHugs: -P../lib:../lib/hugs:../lib/exts +G ffi/FileIO.hs < /dev/null expected stderr not matched by reality *** /tmp/no_stderr27827 Tue Jun 18 21:34:25 2002 --- /tmp/runtest27827.2 Tue Jun 18 21:34:26 2002 *************** *** 0 **** --- 1,3 ---- + MarshalUtils.c: In function `hugsprim_memcpy_0': + MarshalUtils.c:22: warning: conflicting types for built-in function `memcpy' + MarshalUtils.c:22: warning: extern declaration of `memcpy' doesn't match global one -- !!! Testing system calls Revision Changes Path 1.3 +8 -0 hugs98/tests/testScript

1 0

cvs commit: hugs98/tests .cvsignore testScript hugs98/tests/ffi .cvsignore Sin.hs Sin.input Sin.output
by Alastair Reid 18 Jun '02

18 Jun '02

reid 2002/06/18 13:02:08 PDT Modified files: tests .cvsignore testScript Added files: tests/ffi .cvsignore Sin.hs Sin.input Sin.output Log: Simple runtime checks Revision Changes Path 1.2 +1 -1 hugs98/tests/.cvsignore 1.2 +4 -0 hugs98/tests/testScript

1 0

cvs commit: hugs98/tests/ffi types_aux.c
by Alastair Reid 18 Jun '02

18 Jun '02

reid 2002/06/18 12:44:28 PDT Added files: tests/ffi types_aux.c Log: forgeto

1 0

cvs commit: hugs98/src/unix configure configure.in hugs98/tests config.in testScript Readme testScript.in
by Alastair Reid 18 Jun '02

18 Jun '02

reid 2002/06/18 12:37:54 PDT Modified files: src/unix configure configure.in tests Readme Added files: tests config.in testScript Removed files: tests testScript.in Log: Important change to testscript - read this if you are adding/modifying tests. I have tweaked testscript so that it consists of: 1) A small variable part config.in 2) A large fixed part testScript I have removed testScript.in and tweaked configure to modify the small variable part. You no longer have to remember to copy changes you're testing in testScript over into testScript.in. Revision Changes Path 1.22 +16 -12 hugs98/src/unix/configure 1.25 +2 -2 hugs98/src/unix/configure.in 1.6 +3 -9 hugs98/tests/Readme

1 0

cvs commit: hugs98/tests testScript.in hugs98/tests/ffi Forms.hs Loaded.output Types.hs forms_aux.c forms_aux.h
by Alastair Reid 18 Jun '02

18 Jun '02

reid 2002/06/18 12:14:39 PDT Modified files: tests testScript.in Added files: tests/ffi Forms.hs Loaded.output Types.hs forms_aux.c forms_aux.h Log: First cut at a test suite for the ffi. As you'd expect you do 'sh ./testScript ffi' to run the test. These checks only check that typechecking and code generation works ok for correct input. So far, there are no checks for erroneous input and there are no checks that it executes ok. Revision Changes Path 1.23 +15 -3 hugs98/tests/testScript.in

2 1

RE: cvs commit: hugs98/src ffi.c
by Simon Marlow 18 Jun '02

18 Jun '02

> > btw The storable instances are bogus - fixing now > > Ooops, they were in fact fine (I got confused over whether they should > use byte offsets or element offsets). > > But while killing my commit of the 'fixed' code I got myself > in this state: > > cvs server: [09:47:46] waiting for reid's lock in > /home/cvs/root/hugs98/lib/extscvs server: [09:48:16] waiting > for reid's lock in /home/cvs/root/hugs98/lib/extscvs server: > [09:48:46] waiting for reid's lock in > /home/cvs/root/hugs98/lib/extsKilled by signal 2. > > I have no idea how to clean this up. I assume you've probably fixed this by now. For future reference (and anyone else who runs into this problem): stale locks tend to be left around when CVS crashes, and have to be removed by hand by logging on to cvs.haskelll.org with SSH, and removing the appropriate lock file from /home/cvs/root/<wherever> which is usually called something beginning with '.#cvs' IIRC. Cheers, Simon

2 1

cvs commit: hugs98/src ffi.c
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 17:52:14 PDT Modified files: src ffi.c Log: Tweakings to make graphics/tests/GTest work. Made sure that if the Haskell source code asks for a ForeignObj it gets a ForeignObj not the current half-hearted, incomplete, broken attempt at ForeignPtr. Revision Changes Path 1.13 +5 -5 hugs98/src/ffi.c

1 0

cvs commit: hugs98/src ffi.c prelude.h
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 17:41:59 PDT Modified files: src ffi.c prelude.h Log: Don't put quotes round libraries because the libraries are often lists of files. (Well, they are at the moment - discussion on ffi list may change this.) I've left the quotes in place on the windows part of the code but Sigbjorn will probably find he has to remove them. Doubled the amount of Text space allocated. The Xlib ffi code has a lot of strings in it. It may be worth looking for a way to eliminate some of these strings to save a little memory. With these changes and the recent greencard commits, the following makes HGL build and run correctly. green-card --target ffi --include-dir /home/reid/local/green-card/lib/ghc -fX_stub_ffi.c -fcbits/auxiliaries.c -f-L/usr/X11R6/lib -f-lX11 -f-lc X.gc -o X.hs green-card --target ffi --include-dir /home/reid/local/green-card/lib/ghc -fXlib_stub_ffi.c -fcbits/auxiliaries.c -f-L/usr/X11R6/lib -f-lX11 -f-lc Xlib.gc -o Xlib.hs green-card --target ffi --include-dir /home/reid/local/green-card/lib/ghc -fXlib_StdDIS_stub_ffi.c -fcbits/auxiliaries.c -f-L/usr/X11R6/lib -f-lX11 -f-lc Xlib_StdDIS.gc -o Xlib_StdDIS.hs cd ~/src/hugs-working make install make ./hugs -P/home/reid/src/hslibs/xlib:/home/reid/src/hslibs/graphics/lib/x11: \ ~/src/hslibs/graphics/demos/HelloWorld.hs WooHoo! Revision Changes Path 1.12 +3 -3 hugs98/src/ffi.c 1.36 +3 -3 hugs98/src/prelude.h

1 0

cvs commit: hugs98/src static.c
by Sigbjorn Finne 17 Jun '02

17 Jun '02

sof 2002/06/17 17:19:49 PDT Modified files: src static.c Log: checkImportEntity: correctly record imports/hidings of abstract tycons/classes on the computed imports list. Revision Changes Path 1.73 +6 -2 hugs98/src/static.c

1 0

cvs commit: hugs98/tests testScript.in
by Sigbjorn Finne 17 Jun '02

17 Jun '02

sof 2002/06/17 17:16:48 PDT Modified files: tests testScript.in Log: hook up static/mod112.hs Revision Changes Path 1.22 +1 -0 hugs98/tests/testScript.in

1 0

cvs commit: hugs98/tests/static mod112.hs
by Sigbjorn Finne 17 Jun '02

17 Jun '02

sof 2002/06/17 17:16:14 PDT Added files: tests/static mod112.hs Log: regression test for the handling of abstract class 'hiding's.

1 0

cvs commit: hugs98/tests testScript.in
by Sigbjorn Finne 17 Jun '02

17 Jun '02

sof 2002/06/17 17:09:23 PDT Modified files: tests testScript.in Log: factor out common options Revision Changes Path 1.21 +226 -222 hugs98/tests/testScript.in

1 0

cvs commit: hugs98/tests/static mod111.hs
by Sigbjorn Finne 17 Jun '02

17 Jun '02

sof 2002/06/17 17:08:57 PDT Added files: tests/static mod111.hs Log: regression test for the handling of abstract tycon 'hiding's.

1 0

cvs commit: hugs98/lib/exts ForeignObj.hs StablePtr.hs
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 15:57:10 PDT Modified files: lib/exts ForeignObj.hs StablePtr.hs Log: Added new* synonyms for the old make* functions Revision Changes Path 1.2 +3 -0 hugs98/lib/exts/ForeignObj.hs 1.2 +3 -0 hugs98/lib/exts/StablePtr.hs

1 0

cvs commit: hugs98/src ffi.c
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 15:56:15 PDT Modified files: src ffi.c Log: Add Addr support back in to make life with greencard --target=ffi easier - blech Revision Changes Path 1.11 +5 -2 hugs98/src/ffi.c

1 0

cvs commit: hugs98/src HsFFI.h builtin.c connect.h ffi.c static.c
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 14:46:48 PDT Modified files: src HsFFI.h builtin.c connect.h ffi.c static.c Log: Made dynamic and wrapper forms work. This is my current test program: import Foreign import Exception import Prelude hiding (read) tests = do putStrLn "\nTesting sin==mysin (should return lots of Trues)" print (testSin sin mysin) putStrLn "\nTesting errno" err <- peek errno putStrLn $ "errno == " ++ show err putStrLn "\nTesting puts (and withString)" withString0 "Test successful" puts putStrLn "\nTesting peekArray0" s <- withString0 "Test successful" (peekArray0 '\0') putStr s putStrLn "\nTesting open, read and close" s <- testRead "test.hs" 200 putStrLn s putStrLn "\nTesting open, write and close" testWrite "/tmp/test_write" "Test successful" putStrLn "\nTesting sin==dynamic_sin (should return lots of Trues)" print (testSin sin (dyn_sin sin_addr)) putStrLn "\nTesting sin==IO wrapped_sin (should return lots of Trues)" sin_addr2 <- wrapIO (return . sin) print (testSin sin (unsafePerformIO . (dyn_sinIO sin_addr2))) freeHaskellFunPtr sin_addr2 putStrLn "\nTesting sin==Id wrapped_sin (should return lots of Trues)" sin_addr3 <- wrapId sin print (testSin sin (dyn_sin sin_addr3)) freeHaskellFunPtr sin_addr3 putStrLn "\nTesting exit" exit 3 testSin f g = [ (f x == g x) | x <- [0,0.01 .. 1] ] foreign import ccall "sin" mysin :: Double -> Double foreign import ccall "dynamic" dyn_sin :: FunPtr (Double -> Double) -> (Double -> Double) foreign import ccall "dynamic" dyn_sinIO :: FunPtr (Double -> IO Double) -> (Double -> IO Double) foreign import ccall "&sin" sin_addr :: FunPtr (Double -> Double) foreign import ccall "wrapper" wrapId :: (Double -> Double) -> IO (FunPtr (Double -> Double)) foreign import ccall "wrapper" wrapIO :: (Double -> IO Double) -> IO (FunPtr (Double -> IO Double)) foreign import ccall safe "static stdlib.h &errno" errno :: Ptr Int withString s = bracket (newArray s) free withString0 s = bracket (newArray0 '\0' s) free withBuffer sz m = do b <- mallocArray sz sz' <- m b s <- peekArray sz' b free b return s foreign import ccall puts :: Ptr Char -> IO Int foreign import ccall "open" open' :: Ptr Char -> Int -> IO Int foreign import ccall "open" open2' :: Ptr Char -> Int -> Int -> IO Int foreign import ccall "creat" creat' :: Ptr Char -> Int -> IO Int foreign import ccall close :: Int -> IO Int foreign import ccall "read" read' :: Int -> Ptr Char -> Int -> IO Int foreign import ccall "write" write' :: Int -> Ptr Char -> Int -> IO Int creat s m = withString0 s $ \s' -> unix "creat" $ creat' s' m open s m = withString0 s $ \s' -> unix "open" $ open' s' m open2 s m n = withString0 s $ \s' -> unix "open2" $ open2' s' m n write fd s = withString0 s $ \s' -> unix "write" $ write' fd s' (length s) read fd sz = withBuffer sz $ \s' -> unix "read" $ read' fd s' sz unix s m = do x <- m if x < 0 then do err <- peek errno ioError $ userError $ s ++ ": " ++ show (x,err) else return x testRead fn sz = bracket (open fn 0) close (flip read sz) testWrite fn s = bracket (open2 fn (512+64+1) 511) close (flip write s) foreign import ccall exit :: Int -> IO () -- Various bits of rubbish. -- foreign import ccall "static stdlib.h exit" (***) :: Ptr Char -> Ptr Char -> IO Int -- -- foreign import ccall safe "static stdlib.h printf" (+++) :: Ptr Char -> Ptr Char -> IO Int -- foreign import ccall safe "static stdlib.h &errno" illegal_foo :: Ptr Int -- -- foreign import ccall safe "wrapper" illegal_bar :: Char -> IO (FunPtr Char) -- foreign import ccall safe "dynamic" illegal_baz :: FunPtr Char -> Char -- foreign export ccall "id_charstar" id :: Ptr Char -> Ptr Char Revision Changes Path 1.4 +8 -3 hugs98/src/HsFFI.h 1.26 +30 -5 hugs98/src/builtin.c 1.38 +3 -3 hugs98/src/connect.h 1.10 +36 -11 hugs98/src/ffi.c 1.72 +20 -4 hugs98/src/static.c

1 0

cvs commit: hugs98/lib/exts Foreign.hs Ptr.hs FunPtr.hs
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 14:44:52 PDT Modified files: lib/exts Foreign.hs Ptr.hs Removed files: lib/exts FunPtr.hs Log: Killed FunPtr - I'd misread the ffi spec when I added it. Brought FunPtr impln closer to ffi spec. Added unsafePerformIO to Foreign Revision Changes Path 1.5 +2 -0 hugs98/lib/exts/Foreign.hs 1.2 +19 -3 hugs98/lib/exts/Ptr.hs

1 0

cvs commit: hugs98/src storage.c
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 14:43:11 PDT Modified files: src storage.c Log: Added FLOATCELL support to debug printer Revision Changes Path 1.43 +5 -2 hugs98/src/storage.c

1 0

cvs commit: hugs98/lib Prelude.hs
by Alastair Reid 17 Jun '02

17 Jun '02

reid 2002/06/17 14:42:37 PDT Modified files: lib Prelude.hs Log: Added a decent error message when the IO scheduler loop hits an unrecognised thunk. The previous error message of 'undefined' didn't really do it for me. Revision Changes Path 1.27 +1 -1 hugs98/lib/Prelude.hs

1 0

cvs commit: hugs98/src ffi.c
by Sigbjorn Finne 17 Jun '02

17 Jun '02

sof 2002/06/17 09:29:15 PDT Modified files: src ffi.c Log: Simple-minded additions for compiling via MSVC Revision Changes Path 1.9 +20 -4 hugs98/src/ffi.c

2 2