RE: [BUG REPORT] gnulib Android NDK and/or Cygwin build failure regression after 0c8a563f

[Chris Renshaw]

Sorry for taking so long to reply, I had a SSD failure and it's taken this long to get things up and running again.

 

SHELLOPTS='igncr'

 

doesn't work, but:

 

export SHELLOPTS

set -o igncr

 

does, so that is a viable workaround for Cygwin at least.

 

Setting these globally like this can break other things in Cygwin, however, which is why they aren't the default.

 

My proposed patch in my previous messages sanitizes things to work with the default env everywhere, including Cygwin, since the Windows Android NDK will always output CRs whether in Cygwin or not. I believe this still to be the best approach; busybox has adopted something similar for their build scripts to accommodate Cygwin and the Windows Android NDK.

 

Thanks again for your consideration.

 

Chris

 

From: Paul Eggert
Sent: April 24, 2022 5:38 AM
To: osm0sis@outlook.com; bug-gnulib@gnu.org
Subject: Re: [BUG REPORT] gnulib Android NDK and/or Cygwin build failure regression after 0c8a563f

 

On 4/3/22 19:22, osm0sis@outlook.com wrote:
> Examined a copy of the conftest.out file the script parses those from and noticed it had CR+LF due to being from Windows Android NDK and/or on Cygwin,

Sounds like your Cygwin shell is misconfigured. You might try setting
SHELLOPTS='igncr' in your environment before running the shell.