Re: Fink for Ventura / Xcode 15?

FYI: I ran into the same issue with building lzo2

./configure --prefix=/opt/sw  --disable-asm --disable-dependency-tracking --disable-static --enable-shared
configure: Configuring LZO 2.10
checking build system type... x86_64-apple-darwin22.6.0
checking host system type... x86_64-apple-darwin22.6.0
checking target system type... x86_64-apple-darwin22.6.0
checking whether to enable maintainer-specific portions of Makefiles... no
checking for gcc... gcc
checking whether the C compiler works... no
configure: error: in `/opt/sw/src/fink.build/lzo2-2.10-1/lzo-2.10':
configure: error: C compiler cannot create executables
See `config.log' for more details
### execution of ./configure failed, exit code 77


The config.log output is
configure:3151: checking whether the C compiler works
configure:3173: gcc -g -O2 -fstack-protector -Wformat -Werror=format-security -fPIE -D_FORTIFY_SOURCE=2 -I/opt/sw/include -Wl,-read_only_stubs -Wl,-bind_at_load -fPIE -Wl,-pie -L/opt/sw/lib/x86_64-darwin -L/opt/sw/lib -Wl,-headerpad_max_install_names conftest.c  >&5
ld: unknown options: -read_only_stubs 
clang: error: linker command failed with exit code 1 (use -v to see invocation)
configure:3177: $? = 1
configure:3215: result: no

So need that same patch to configure?  or configure.ac?  Or aclocal.m4, not sure how autoconf works that has been as big a mystery as JCL was on the IBM 360 system.

-Scott
> On Sep 24, 2023, at 7:26 AM, Hanspeter Niederstrasser <[email protected]> wrote:
> 
> On 9/23/23 5:27 PM, Scott Hannahs via Fink-users wrote:
>>> On Sep 23, 2023, at 4:43 PM, Hanspeter Niederstrasser <[email protected]> wrote:
>>> 
>>> You want a part earlier in config.log. Search for the string "checking whether the C compiler works" and then read down for the actual error that led configure to declare "C compiler cannot create executables".
>>> 
>>> Also, there's been a couple updates to deal with macOS 13.6 and 12.7 in git, so make sure you have git checkout that's up to date (via git pull).
>>> 
>>> Hanspeter
>>> 
>>> -- 
>>> Ad Astra
>> Thanks, didn’t see that.  So one of the ld options is no longer valid?
>> configure:3435: checking whether the C compiler works
>> configure:3457: gcc -g -O2 -fstack-protector -Wformat -fPIE -Wno-unused-parameter -Wno-missing-field-initializers -Wno-cast-align -Wno-format-security -D_FORTIFY_SOURCE=2 -I/opt/sw/boots
>> trap/include -Wl,-read_only_stubs -Wl,-bind_at_load -fPIE -Wl,-pie -L/opt/sw/bootstrap/lib/0 -L/opt/sw/bootstrap/lib -Wl,-headerpad_max_install_names conftest.c >&5
>> ld: unknown options: -read_only_stubs
>> clang: error: linker command failed with exit code 1 (use -v to see invocation)
>> configure:3461: $? = 1
>> configure:3499: result: no
>> That option is in the man page.
>> -read_only_stubs
>>              [i386 only] Makes the __IMPORT segment of a final linked images read-only.  This option makes a program slightly more secure in that the JMP instructions in the i386 fast
>>              stubs cannot be easily overwritten by malicious code.  The downside is the dyld must use mprotect() to temporarily make the segment writable while it is binding the stubs.
>> And this is building on an Intel system that “arch” says is i386.
> 
> Can you apply the attached patch from the directory where you run ./bootstrap ?
> 
> `patch -p1 /path/to/patchfile`
> 
> This gets rid of the now problematic -read_only_stubs flag. Once applied, run `./bootstrap` again.
> 
> Hanspeter
> <no-read-only-stubs.patch>

_______________________________________________
Fink-users mailing list
[email protected]
List archive:
https://sourceforge.net/p/fink/mailman/fink-users/
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-users