This is the mail archive of the cygwin-apps@sourceware.cygnus.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [cet@carlthompson.net: Re: Extreme newbie here...]


On Fri, Jun 09, 2000 at 12:38:54PM -0500, Parker, Ron wrote:
>> >bash-2.04$ g++ -mno-cygwin -fno-rtti hello.cpp -o hello
>> >/usr/lib/gcc-lib/i686-pc-cygwin/2.95.2/libgcc.a(_eh.o): In function
>> >`_eh_rtime_match':
>> >/home/src/gcc-2.95.2/gcc/./libgcc2.c(.text+0x389): undefined 
>> reference to
>> >`_impure_ptr'
>> >collect2: ld returned 1 exit status
>> >
>> >If your source code will compile with -fno-exceptions, it 
>> should work in
>> >most cases.
>> 
>> Hmm.  We noticed this when building strace and Mumit 
>> submitted a workaround
>> for it.  If you search the cygwin-developers archive you 
>> might find the
>> reason for this.  I think that the gcc library is somehow 
>> pulling in bits
>> of the cygwin library.
>
>Unfortunately compiling with -fno-exceptions isn't always an option and
>adding a dummy "int _impure_ptr;" declaration as was done in strace does not
>work always work.  Even the simple eh_hello.cpp I wrote up will GP if you
>try this.  I have also compiled and linked against Mumit's ming-extras
>package it still dies.

Oh, I didn't mean to propose this as a fix.  It definitely isn't.
"We" really should fix this problem.

>I feel like the great white bug hunter.  "Dr. Livingston, I presume?"  
>
>Which brings the following sick story to mind,
>http://www.qnx.com/~glen/deadbeef/3821.html.  It is gross and disgusting.
>You have been warned.

I have seen this story before.  It is quite... tragic.

cgf

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]