[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Bug-apl] [bug] problem with simple code
From: |
Alexey Veretennikov |
Subject: |
Re: [Bug-apl] [bug] problem with simple code |
Date: |
Mon, 29 Feb 2016 18:58:49 +0100 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/25.0.91 (darwin) |
Hi,
Thanks, I can confirm what it is fixed, works for me now.
Juergen Sauermann <address@hidden> writes:
> Hi Alexey,
>
> I could reproduce the fault now, and I believe it is fixed in SVN 698.
>
> The warning of your compiler is somewhat bogus and MAY be fixed as well.
>
> /// Jürgen
>
> On 02/28/2016 06:26 PM, Alexey Veretennikov wrote:
>
>
> Hi,
>
> I still got the problem with compilation with "make develop":
> -----
> Command.cc: In static member function 'static bool
> Command::is_directory(dirent*, const UTF8_string&)':
> Command.cc:923:1: error: assuming signed overflow does not occur when
> assuming that (X + c) < X is always false [-Werror=strict-overflow]
> Command::is_directory(dirent * entry, const UTF8_string & path)
> ^
> In file included from UCS_string.hh:28:0,
> from Error.hh:27,
> from Shape.hh:25,
> from PrintBuffer.hh:27,
> from Cell.hh:27,
> from CharCell.hh:24,
> from Command.cc:27:
> Simple_string.hh:158:9: error: assuming signed overflow does not occur when
> assuming that (X + c) < X is always false [-Werror=strict-overflow]
> if (items_valid >= items_allocated) extend();
> ^
> Simple_string.hh:158:9: error: assuming signed overflow does not occur when
> assuming that (X + c) < X is always false [-Werror=strict-overflow]
> mv -f .deps/apl-Error.Tpo .deps/apl-Error.Po
> -----
>
> And the version built with just "make" still has the same bug. (I just
> do
> )load bug1
> clean triples 5
> ==============================================================================
> Assertion failed: 0
> in Function: init
> in file: Cell.cc:48
>
> Call stack:
>
> ----------------------------------------
> -- Stack trace at Cell.cc:48
> ----------------------------------------
> ========================================
>
> SI stack:
>
> Depth: 3
> Exec: 0x7f99d5103c00
> Safe ex: no
> Pmode: ∇ clean[1]
> PC: 12 ENDL
> Stat: λ← ∪Sort¨ ({⍵≡∪⍵}¨⍵)/⍵
> err_code: 0x0
> thrown: at StateIndicator.cc:40
> e_msg_1: 'No Error'
> e_msg_2: ''
> e_msg_3: ''
>
> Depth: 2
> Exec: 0x7f99d36ac650
> Safe ex: no
> Pmode: ◊ clean triples 5
> PC: 4 RETURN_STATS
> Stat: clean triples 5
> err_code: 0x0
> thrown: at StateIndicator.cc:40
> e_msg_1: 'No Error'
> e_msg_2: ''
> e_msg_3: ''
>
> Depth: 1
> Exec: 0x7f99d5104e00
> Safe ex: no
> Pmode: ◊ a
> PC: 0
> ==============================================================================
> Assertion failed: idx < items_valid
> in Function: operator[]
> in file: Simple_string.hh:140
>
> Call stack:
> *** do_Assert() called recursively ***
> ==============================================================================
>
>
> Juergen Sauermann <address@hidden> writes:
>
>
> Hi Alexey,
>
> thanks, fixed in SVN 697.
>
> /// Jürgen
>
> On 02/27/2016 10:29 PM, Alexey Veretennikov wrote:
>
>
> Thanks, now it is better and I was able to start compilation with "make
> develop". However I quickly stuck with the compilation error:
>
> In file included from Common.hh:53:0,
> from Backtrace.hh:26,
> from Error.hh:24,
> from IndexIterator.cc:21:
> IndexIterator.hh: In member function 'virtual ShapeItem
> TrueIndexIterator::get_value() const':
> IndexIterator.hh:112:20: error: comparison between signed and unsigned
> integer expressions [-Werror=sign-compare]
> { Assert(pos < indices.size()); return indices[pos]; }
> ^
> Assert.hh:50:26: note: in definition of macro 'Assert'
> #define Assert(x) if (!(x)) do_Assert(#x, __FUNCTION__, __FILE__,
> __LINE__)
> ^
> IndexIterator.hh: In member function 'virtual ShapeItem
> TrueIndexIterator::get_pos(ShapeItem) const':
> IndexIterator.hh:116:18: error: comparison between signed and unsigned
> integer expressions [-Werror=sign-compare]
> { Assert(i < indices.size()); return indices[i]; }
> ^
> Assert.hh:50:26: note: in definition of macro 'Assert'
> #define Assert(x) if (!(x)) do_Assert(#x, __FUNCTION__, __FILE__,
> __LINE__)
> ^
> mv -f .deps/apl-IndexExpr.Tpo .deps/apl-IndexExpr.Po
> g++ -DHAVE_CONFIG_H -I. -I.. -Werror -Wall -Wno-strict-aliasing -g -O2
> -MT apl-IntCell.o -MD -MP -MF .deps/apl-IntCell.Tpo -c -o apl-IntCell.o `test
> -f 'IntCell.cc' || echo './'`IntCell.cc
> cc1plus: all warnings being treated as errors
>
> Juergen Sauermann <address@hidden> writes:
>
> Hi Alexey,
>
> make clean
>
> before make
>
> has the same effect. If you have the autotools installed (looks like you
> have) then the
> following should work as well:
>
> autoreconf
> ./configure
> make develop
>
> /// Jürgen
>
> On 02/27/2016 07:03 PM, Alexey Veretennikov wrote:
>
>
> Hi,
>
> I've just removed local version and checked out again. Did ./configure
> and make, and the bug still there.
> I cannot build with "make develop" since it complains about absense of
> aclocal-1.14 (I have aclocal installed).
>
> However I'm compiling and running on OSX 10.10.5
>
> Juergen Sauermann <address@hidden> writes:
>
> Hi Alexey,
>
> Hm, it works on my machine. Did you recently upgrade from SVN?
> It can happen that changes in header files are not detected if ./configure is
> called without arguments
> and then the sources are uodated from SVN, Please try the following:
>
> make clean develop
>
> and see if the problem remains.
>
> /// Jürgen
>
> On 02/27/2016 12:26 AM, Alexey Veretennikov wrote:
>
>
> Hi,
>
> I'm studying the paper http://archive.vector.org.uk/art10501380
>
> Here defined 3 functions:
> ⍝ simple sort
> Sort←{⍵[⍋⍵]}
>
> R←∇triples N;all;i
> ⍝ All possible triples of numbers 1..9
> i←⍳9
> all←,i∘.,i∘.,i
> R←(N=+/¨all)/all
>
> ⍝ remove those triples which contain repeating entries or not unique
> clean←{ ∪Sort¨ ({⍵≡∪⍵}¨⍵)/⍵ }
>
> And then try to run
> clean triples 4
> lead to crash(assertion).
> Version APL: 1.5 / 696,
> Workspace attached.
>
> The assert log below:
>
> ==============================================================================
> Assertion failed: 0
> in Function: init
> in file: Cell.cc:48
>
> Call stack:
>
> ----------------------------------------
> -- Stack trace at Cell.cc:48
> ----------------------------------------
> 0x0 @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> ========================================
>
> SI stack:
>
> Depth: 4
> Exec: 0x7fe663500dd0
> Safe ex: no
> Pmode: ∇ clean[1]
> PC: 12 ENDL
> Stat:
>
> ====================================================
> SEGMENTATION FAULT
>
> ----------------------------------------
> -- Stack trace at main.cc:63
> ----------------------------------------
> 0x0 @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0xa @@@@
> 0x0 @@@@
> 0x0 @@@@
> 0xa @@@@
> ========================================
> ====================================================
>
> Goodbye.
>
> Process apl exited abnormally with code 3
>
>
>
>
>
>
>
>
>
>
--
Br,
/Alexey
- [Bug-apl] [bug] problem with simple code, Alexey Veretennikov, 2016/02/26
- Re: [Bug-apl] [bug] problem with simple code, Juergen Sauermann, 2016/02/27
- Re: [Bug-apl] [bug] problem with simple code, Alexey Veretennikov, 2016/02/27
- Re: [Bug-apl] [bug] problem with simple code, Juergen Sauermann, 2016/02/27
- Re: [Bug-apl] [bug] problem with simple code, Alexey Veretennikov, 2016/02/27
- Re: [Bug-apl] [bug] problem with simple code, Juergen Sauermann, 2016/02/28
- Re: [Bug-apl] [bug] problem with simple code, Alexey Veretennikov, 2016/02/28
- Re: [Bug-apl] [bug] problem with simple code, Juergen Sauermann, 2016/02/29
- Re: [Bug-apl] [bug] problem with simple code,
Alexey Veretennikov <=