September 2014 - ghc-tickets

Re: [GHC] #6086: Cross compilation fails using system linker for other architecture binaries
by GHC 10 Sep '14

10 Sep '14

#6086: Cross compilation fails using system linker for other architecture binaries ----------------------------------------------+--------------------------- Reporter: mtjm | Owner: Type: bug | Status: new Priority: high | Milestone: 7.8.4 Component: Build System | Version: 7.8.3 Resolution: | Keywords: Operating System: Linux | Architecture: arm Type of failure: Building GHC failed | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | ----------------------------------------------+--------------------------- Comment (by MikolajKonarski): I get the original problem (the need to manually insert --ld- option="-melf_i386") already when using a 32bit GHC on 64bit Linux (no proper cross-compilation): https://ghc.haskell.org/trac/ghc/ticket/9421#comment:2 It would be helpful if that and other parameters where inserted automatically based on HOST/BUILD/TARGET. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6086#comment:13> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6086: Cross compilation fails using system linker for other architecture binaries
by GHC 10 Sep '14

10 Sep '14

#6086: Cross compilation fails using system linker for other architecture binaries ----------------------------------------------+--------------------------- Reporter: mtjm | Owner: Type: bug | Status: new Priority: high | Milestone: 7.8.4 Component: Build System | Version: 7.8.3 Resolution: | Keywords: Operating System: Linux | Architecture: arm Type of failure: Building GHC failed | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | ----------------------------------------------+--------------------------- Comment (by rwbarton): libgmp should be being configured with `--host=arm-unknown-linux-gnueabi`. I think this is a bug in the build system and currently you probably just want to use integer-simple instead. I don't think `make quick-cross` actually does anything... you should set up a `build.mk` (copied from `build.mk.sample`) and add `INTEGER_LIBRARY = integer-simple`. I think pretty much any `BuildFlavour` should work but note that the `quick-cross` and `perf-cross` flavours only build a ''cross-compiler'' and not a ''cross-compiled compiler'' that runs on the target. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6086#comment:12> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6086: Cross compilation fails using system linker for other architecture binaries
by GHC 10 Sep '14

10 Sep '14

#6086: Cross compilation fails using system linker for other architecture binaries ----------------------------------------------+--------------------------- Reporter: mtjm | Owner: Type: bug | Status: new Priority: high | Milestone: 7.8.4 Component: Build System | Version: 7.8.3 Resolution: | Keywords: Operating System: Linux | Architecture: arm Type of failure: Building GHC failed | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | ----------------------------------------------+--------------------------- Comment (by cetu86): I did the maintainer-clean, then a {{{ perl boot }}} and then again {{{ ./configure --target=arm-unknown-linux-gnueabi --enable-unregisterised --with-gcc=`which arm-unknown-linux-gnueabi-gcc` --with-nm=/usr/bin/nm --prefix=/home/daniel/haskell/ghc-build && make quick-cross }}} Now I get {{{ mv libraries/integer-gmp/gmp/gmpbuild/ltmain.sh libraries/integer- gmp/gmp/gmpbuild/ltmain.sh.orig sed 's#cmd //c echo "\$1"#cmd /c "echo $1"#' < libraries/integer- gmp/gmp/gmpbuild/ltmain.sh.orig > libraries/integer- gmp/gmp/gmpbuild/ltmain.sh cd libraries/integer-gmp/gmp; (set -o igncr 2>/dev/null) && set -o igncr; export SHELLOPTS; \ PATH=`pwd`:$PATH; \ export PATH; \ cd gmpbuild && \ CC=/home/daniel/x-tools/arm-unknown-linux-gnueabi/bin/arm-unknown- linux-gnueabi-gcc NM=/usr/bin/nm AR=/home/daniel/x-tools/arm-unknown- linux-gnueabi/bin/arm-unknown-linux-gnueabi-ar /bin/sh ./configure \ --enable-shared=no \ --host=x86_64-unknown-linux --build=x86_64-unknown-linux checking build system type... x86_64-unknown-linux-gnu checking host system type... x86_64-unknown-linux-gnu checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes checking for a thread-safe mkdir -p... /usr/bin/mkdir -p checking for gawk... gawk checking whether make sets $(MAKE)... yes checking whether to enable maintainer-specific portions of Makefiles... no checking ABI=64 checking whether /home/daniel/x-tools/arm-unknown-linux-gnueabi/bin/arm- unknown-linux-gnueabi-gcc is gcc... yes checking compiler /home/daniel/x-tools/arm-unknown-linux-gnueabi/bin/arm- unknown-linux-gnueabi-gcc -O2 -pedantic -m64 ... no checking ABI=32 checking whether /home/daniel/x-tools/arm-unknown-linux-gnueabi/bin/arm- unknown-linux-gnueabi-gcc is gcc... yes checking compiler /home/daniel/x-tools/arm-unknown-linux-gnueabi/bin/arm- unknown-linux-gnueabi-gcc -m32 -O2 -pedantic -fomit-frame-pointer ... no checking compiler /home/daniel/x-tools/arm-unknown-linux-gnueabi/bin/arm- unknown-linux-gnueabi-gcc -O2 -pedantic -fomit-frame-pointer ... no, program does not run configure: error: could not find a working compiler, see config.log for details make[1]: *** [libraries/integer-gmp/gmp/gmp.h] Error 1 make: *** [quick-cross] Error 2 }}} Again, it's seems like it's trying to compile gmp with an arm compiler for x64, or am I wrong? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6086#comment:11> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6086: Cross compilation fails using system linker for other architecture binaries
by GHC 10 Sep '14

10 Sep '14

#6086: Cross compilation fails using system linker for other architecture binaries ----------------------------------------------+--------------------------- Reporter: mtjm | Owner: Type: bug | Status: new Priority: high | Milestone: 7.8.4 Component: Build System | Version: 7.8.3 Resolution: | Keywords: Operating System: Linux | Architecture: arm Type of failure: Building GHC failed | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | ----------------------------------------------+--------------------------- Comment (by rwbarton): Hmm, so that's in libffi. Check how it was configured (`libffi/build/arm- unknown-linux-gnueabi/config.log` should tell you) and see whether anything looks amiss there. Perhaps you needed to clean more before re- configuring (`make distclean` should probably be enough, but `make maintainer-clean` is the ultimate hammer). -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6086#comment:10> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6086: Cross compilation fails using system linker for other architecture binaries
by GHC 10 Sep '14

10 Sep '14

#6086: Cross compilation fails using system linker for other architecture binaries ----------------------------------------------+--------------------------- Reporter: mtjm | Owner: Type: bug | Status: new Priority: high | Milestone: 7.8.4 Component: Build System | Version: 7.8.3 Resolution: | Keywords: Operating System: Linux | Architecture: arm Type of failure: Building GHC failed | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | ----------------------------------------------+--------------------------- Comment (by cetu86): Thanks for that quick answer. I did as you told me, also passing the argument {{{ --with-gcc=`which arm-unknown-linux-gnueabi-gcc` }}} to configure. Now it goes past the last error. But some way down the road this happens {{{ /bin/sh ./libtool --mode=compile /usr/bin/gcc -DHAVE_CONFIG_H -I. -I.. -I. -I../include -Iinclude -I../src -DFFI_BUILDING -I. -I../include -Iinclude -I../src -DFFI_BUILDING -g -fno-stack-protector -w -MT src/arm/sysv.lo -MMD -MP -MF $depbase.Tpo -c -o src/arm/sysv.lo ../src/arm/sysv.S &&\ mv -f $depbase.Tpo $depbase.Plo libtool: compile: /usr/bin/gcc -DHAVE_CONFIG_H -I. -I.. -I. -I../include -Iinclude -I../src -DFFI_BUILDING -I. -I../include -Iinclude -I../src -DFFI_BUILDING -g -fno-stack-protector -w -MT src/arm/sysv.lo -MMD -MP -MF src/arm/.deps/sysv.Tpo -c ../src/arm/sysv.S -fPIC -DPIC -o src/arm/.libs/sysv.o ../src/arm/sysv.S: Assembler messages: ../src/arm/sysv.S:64: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:160: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:161: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:162: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:163: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:164: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:166: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:167: Error: unknown pseudo-op: `.arm' ../src/arm/sysv.S:167: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:168: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:169: Error: no such instruction: `stmfd sp!,{r0-r3,fp,lr}' ../src/arm/sysv.S:170: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:171: Error: too many memory references for `mov' ../src/arm/sysv.S:173: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:175: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:176: Error: too many memory references for `sub' ../src/arm/sysv.S:178: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:179: Error: too many memory references for `mov' ../src/arm/sysv.S:180: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:182: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:183: Error: no such instruction: `bl ffi_prep_args' ../src/arm/sysv.S:185: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:186: Error: no such instruction: `ldmia sp,{r0-r3}' ../src/arm/sysv.S:188: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:189: Error: too many memory references for `sub' ../src/arm/sysv.S:190: Error: no such instruction: `ldr ip,[fp]@load fn()in advance' ../src/arm/sysv.S:191: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:192: Error: no such instruction: `movhs lr,' ../src/arm/sysv.S:193: Error: too many memory references for `add' ../src/arm/sysv.S:195: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:196: Error: too many memory references for `mov' ../src/arm/sysv.S:196: Error: too many memory references for `mov' ../src/arm/sysv.S:198: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:199: Error: too many memory references for `mov' ../src/arm/sysv.S:201: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:202: Error: no such instruction: `ldr r2,[sp,' ../src/arm/sysv.S:204: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:205: Error: no such instruction: `ldr r3,[sp,' ../src/arm/sysv.S:207: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:208: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:209: Error: no such instruction: `beq .Lepilogue' ../src/arm/sysv.S:211: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:212: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:216: Error: no such instruction: `streq r0,[r2]' ../src/arm/sysv.S:217: Error: no such instruction: `beq .Lepilogue' ../src/arm/sysv.S:219: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:220: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:224: Error: no such instruction: `stmeqia r2,{r0,r1}' ../src/arm/sysv.S:227: Error: no such instruction: `beq .Lepilogue' ../src/arm/sysv.S:229: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:230: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:231: Error: no such instruction: `stfeqs f0,[r2]' ../src/arm/sysv.S:232: Error: no such instruction: `beq .Lepilogue' ../src/arm/sysv.S:234: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:235: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:236: Error: no such instruction: `stfeqd f0,[r2]' ../src/arm/sysv.S:244: Error: no such instruction: `ldmia sp!,{r0-r3,fp,pc}' ../src/arm/sysv.S:248: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:269: Error: unknown pseudo-op: `.arm' ../src/arm/sysv.S:269: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:270: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:271: Error: too many memory references for `add' ../src/arm/sysv.S:272: Error: no such instruction: `stmfd sp!,{ip,lr}' ../src/arm/sysv.S:273: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:274: Error: too many memory references for `add' ../src/arm/sysv.S:275: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:276: Error: too many memory references for `sub' ../src/arm/sysv.S:277: Error: invalid char '[' beginning operand 2 `[sp' ../src/arm/sysv.S:278: Error: too many memory references for `add' ../src/arm/sysv.S:279: Error: no such instruction: `bl ffi_closure_SYSV_inner' ../src/arm/sysv.S:280: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:281: Error: no such instruction: `beq .Lretint' ../src/arm/sysv.S:283: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:287: Error: no such instruction: `beq .Lretfloat' ../src/arm/sysv.S:290: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:294: Error: no such instruction: `beq .Lretdouble' ../src/arm/sysv.S:297: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:301: Error: no such instruction: `beq .Lretlongdouble' ../src/arm/sysv.S:304: Error: expecting operand after ','; got nothing ../src/arm/sysv.S:305: Error: no such instruction: `beq .Lretlonglong' ../src/arm/sysv.S:307: Error: too many memory references for `add' ../src/arm/sysv.S:308: Error: no such instruction: `ldmfd sp,{sp,pc}' ../src/arm/sysv.S:310: Error: no such instruction: `ldr r0,[sp]' ../src/arm/sysv.S:311: Error: no such instruction: `b .Lclosure_epilogue' ../src/arm/sysv.S:313: Error: no such instruction: `ldr r0,[sp]' ../src/arm/sysv.S:314: Error: no such instruction: `ldr r1,[sp,' ../src/arm/sysv.S:315: Error: no such instruction: `b .Lclosure_epilogue' ../src/arm/sysv.S:319: Error: no such instruction: `ldfs f0,[sp]' ../src/arm/sysv.S:320: Error: no such instruction: `b .Lclosure_epilogue' ../src/arm/sysv.S:322: Error: no such instruction: `ldfd f0,[sp]' ../src/arm/sysv.S:323: Error: no such instruction: `b .Lclosure_epilogue' ../src/arm/sysv.S:325: Error: no such instruction: `ldfd f0,[sp]' ../src/arm/sysv.S:326: Error: no such instruction: `b .Lclosure_epilogue' ../src/arm/sysv.S:330: Error: junk at end of line, first unrecognized character is `@' ../src/arm/sysv.S:499: Error: no such instruction: `stmfd sp!,{r0-r3}' ../src/arm/sysv.S:500: Error: no such instruction: `ldr r0,[pc]' ../src/arm/sysv.S:501: Error: no such instruction: `ldr pc,[pc]' make[5]: *** [src/arm/sysv.lo] Error 1 make[5]: Leaving directory `/mnt/hd/extra/haskell/ghc/libffi/build/arm- unknown-linux-gnueabi' make[4]: *** [all-recursive] Error 1 make[4]: Leaving directory `/mnt/hd/extra/haskell/ghc/libffi/build/arm- unknown-linux-gnueabi' make[3]: *** [all] Error 2 make[3]: Leaving directory `/mnt/hd/extra/haskell/ghc/libffi/build/arm- unknown-linux-gnueabi' make[2]: *** [all-all] Error 2 make[1]: *** [libffi/stamp.ffi.static-shared.build] Error 2 make: *** [quick-cross] Error 2 }}} Looks like its trying to compile arm-stuff using my normal gcc, not the cross-gcc. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6086#comment:9> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6018: Injective type families
by GHC 10 Sep '14

10 Sep '14

#6018: Injective type families -------------------------------------+------------------------------------- Reporter: lunaris | Owner: jstolarek Type: feature | Status: new request | Milestone: 7.10.1 Priority: normal | Version: 7.4.1 Component: Compiler | Keywords: TypeFamilies, Resolution: | Injective Operating System: | Architecture: Unknown/Multiple Unknown/Multiple | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: #4259 Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by goldfire): Replying to [comment:53 jstolarek]: > Also, if I understood correctly it is possible to have both injectivity and head-injectivity in the compiler, right? Yes, but confusing, especially if there is syntax for both. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6018#comment:55> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6018: Injective type families
by GHC 10 Sep '14

10 Sep '14

#6018: Injective type families -------------------------------------+------------------------------------- Reporter: lunaris | Owner: jstolarek Type: feature | Status: new request | Milestone: 7.10.1 Priority: normal | Version: 7.4.1 Component: Compiler | Keywords: TypeFamilies, Resolution: | Injective Operating System: | Architecture: Unknown/Multiple Unknown/Multiple | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: #4259 Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by jstolarek): Replying to [comment:51 goldfire]: > Replying to [comment:49 jstolarek]: > > I don't think that restricting the use of `result` word in type family head is a big problem - we've already done that with `role`. > > The use of `role` doesn't steal syntax: it can be used only after the word `type` in a top-level declaration, where a capitalized identifier is expected. Oops, sorry for defamation then ;-) Still, I think that the price to pay for using `result` will not be that high. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6018#comment:54> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6018: Injective type families
by GHC 10 Sep '14

10 Sep '14

#6018: Injective type families -------------------------------------+------------------------------------- Reporter: lunaris | Owner: jstolarek Type: feature | Status: new request | Milestone: 7.10.1 Priority: normal | Version: 7.4.1 Component: Compiler | Keywords: TypeFamilies, Resolution: | Injective Operating System: | Architecture: Unknown/Multiple Unknown/Multiple | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: #4259 Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by jstolarek): Replying to [comment:52 goldfire]: > Here, "head" means the head of the RHS type, not the type declaration -- it's the root of the type's AST. Does that help? Yes, now I fully understand. Also, if I understood correctly it is possible to have both injectivity and head-injectivity in the compiler, right? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6018#comment:53> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6018: Injective type families
by GHC 10 Sep '14

10 Sep '14

#6018: Injective type families -------------------------------------+------------------------------------- Reporter: lunaris | Owner: jstolarek Type: feature | Status: new request | Milestone: 7.10.1 Priority: normal | Version: 7.4.1 Component: Compiler | Keywords: TypeFamilies, Resolution: | Injective Operating System: | Architecture: Unknown/Multiple Unknown/Multiple | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: #4259 Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by goldfire): Replying to [comment:50 jstolarek]: > Could you explain what you mean by `RHSs with different heads` and `below the heads`? Up till now I thought that the term "head" refers to `type family F a` part of the declaration. Here, "head" means the head of the RHS type, not the type declaration -- it's the root of the type's AST. Does that help? > > I also don't understand why `G` is not head-injective, but perhaps this will become clear to me once earlier questions are answered. `G` isn't head-injective because both RHSs are headed by `Just`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6018#comment:52> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6018: Injective type families
by GHC 10 Sep '14

10 Sep '14

#6018: Injective type families -------------------------------------+------------------------------------- Reporter: lunaris | Owner: jstolarek Type: feature | Status: new request | Milestone: 7.10.1 Priority: normal | Version: 7.4.1 Component: Compiler | Keywords: TypeFamilies, Resolution: | Injective Operating System: | Architecture: Unknown/Multiple Unknown/Multiple | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: #4259 Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by goldfire): Replying to [comment:49 jstolarek]: > I don't think that restricting the use of `result` word in type family head is a big problem - we've already done that with `role`. The use of `role` doesn't steal syntax: it can be used only after the word `type` in a top-level declaration, where a capitalized identifier is expected. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6018#comment:51> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0