This is the mail archive of the cygwin@cygwin.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]
Other format: [Raw text]

re: overcoming make in openssl


On Wed, 2 Jul 2003, james_lankton wrote:

> Here are the details as requested
>
> I have configured open ssl-0.9.7b with the command ./
> but am having a problem with the make command.
>
>
> Here is the result.
>
>
>
> Administrator@beta ~
> $ cd openssl-0.9.7b
>
> Administrator@beta ~/openssl-0.9.7b
> $ make
> making all in crypto...
> make[1]: Entering directory `/home/Administrator/openssl-0.9.7b/crypto'
> gcc -I. -I.. -I../include -DOPENSSL_SYSNAME_CYGWIN32 -DOPENSSL_THREADS  -DDS
> O_WI
> N32 -DOPENSSL_NO_KRB5 -DTERMIOS -DL_ENDIAN -fomit-frame-pointer -O3 -march=i
> 486
> -Wall -DSHA1_ASM -DMD5_ASM -DRMD160_ASM   -c -o cryptlib.o cryptlib.c
> cryptlib.c:70: parse error before "IMPLEMENT_STACK_OF"
> cryptlib.c:73: warning: return type defaults to `int'
> cryptlib.c: In function `IMPLEMENT_STACK_OF':
> cryptlib.c:73: `CRYPTO_NUM_LOCKS' undeclared (first use in this function)
> cryptlib.c:73: (Each undeclared identifier is reported only once
> cryptlib.c:73: for each function it appears in.)
> cryptlib.c:73: storage class specified for parameter `lock_names'
> cryptlib.c:73: parameter `lock_names' is initialized
> cryptlib.c:76: warning: excess elements in scalar initializer
> cryptlib.c:76: warning: (near initialization for `lock_names')
> cryptlib.c:77: warning: excess elements in scalar initializer
> cryptlib.c:77: warning: (near initialization for `lock_names')
> cryptlib.c:78: warning: excess elements in scalar initializer
> cryptlib.c:78: warning: (near initialization for `lock_names')
> cryptlib.c:79: warning: excess elements in scalar initializer
> cryptlib.c:79: warning: (near initialization for `lock_names')
> cryptlib.c:80: warning: excess elements in scalar initializer
> cryptlib.c:80: warning: (near initialization for `lock_names')
> [...]

I had a similar error this morning when I accidentially declared an array of
structs as a single struct. By the looks of it a header file isn't being
included.

> cryptlib.c:109:3: #error "Inconsistency between crypto.h and cryptlib.c"
> cryptlib.c:115: warning: type defaults to `int' in declaration of `STACK'
> cryptlib.c:115: storage class specified for parameter `STACK'
> cryptlib.c:115: parse error before '*' token
> cryptlib.c:115: declaration for parameter `STACK' but no such parameter
> cryptlib.c:73: declaration for parameter `lock_names' but no such parameter
> cryptlib.c:148: `app_locks' undeclared (first use in this function)
> cryptlib.c:148: warning: implicit declaration of function `sk_new_null'
> cryptlib.c:150: warning: implicit declaration of function `CRYPTOerr'
> cryptlib.c:150: `CRYPTO_F_CRYPTO_GET_NEW_LOCKID' undeclared (first use in
> this f
> unction)
> cryptlib.c:150: `ERR_R_MALLOC_FAILURE' undeclared (first use in this
> function)
> cryptlib.c:153: warning: implicit declaration of function `BUF_strdup'
> cryptlib.c:153: `name' undeclared (first use in this function)
> cryptlib.c:153: warning: assignment makes pointer from integer without a
> cast
> cryptlib.c:158: warning: implicit declaration of function `sk_push'
> cryptlib.c:160: warning: implicit declaration of function `OPENSSL_free'
> cryptlib.c: In function `CRYPTO_num_locks':
> cryptlib.c:168: `CRYPTO_NUM_LOCKS' undeclared (first use in this function)
> [...]

As I think Corinna has already mentioned, there is a build script
(CYGWIN-PATCHES/build.sh ?) that Corinna uses when building the distribution.
Have you tried running that?

Elfyn
-- 


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/


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