On 2022-10-06 13:24, Adam Dinwoodie wrote:
"no iconv" concerns me; I'm not desperately familiar with how iconv
works, but I believe that'll potentially cause issues for rsync users
who aren't using ASCII.  I'd guess the issue is your build environment
is missing a relevant build-time dependency, probably libiconv or
libiconv-devel.


Yeah, come to think of it that one also stuck out to me after I sent the mail. I've updated my cygport and associated artifacts to include those dependencies:
https://github.com/crd477/cygports/tree/main/rsync

Here's an update diff of the features and capabilities, reformatted and sorted for easier comparison:
crd@x13:~/src/cygports/rsync$ diff -b old.txt new.txt
1,2c1,2
< rsync  version 3.2.4dev  protocol version 31
< Copyright (C) 1996-2020 by Andrew Tridgell, Wayne Davison, and others.
---
> rsync  version 3.2.6  protocol version 31
> Copyright (C) 1996-2022 by Andrew Tridgell, Wayne Davison, and others.
13a14,15
>  crtimes
>  hardlink-symlinks
17d18
<  no crtimes
19c20
<  optional protect-args
---
>  optional secluded-args
27,28c28,30
<  no SIMD
<  asm
---
>  no SIMD-roll
>  no asm-MD5
>  no asm-roll

I think the diffs after "19c20" above should be safe to ignore as the release notes explain them: " - Renamed configure's `--enable-simd` option to `--enable-roll-simd` and added the option `--enable-roll-asm` to use the new asm version of the code. Both
   are x86_64/amd64 only.

 - Renamed configure's `--enable-asm` option to `--enable-md5-asm` to avoid
confusion with the asm option for the rolling checksum. It is also honored even when openssl crypto is in use. This allows: normal MD4 & MD5, normal MD4 + asm MD5, openssl MD4 & MD5, or openssl MD4 + asm MD5 depending on the
   configure options selected.

- Made SIMD & asm configure checks default to "no" on non-Linux hosts due to
   various reports of problems on NetBSD & macOS hosts.  These were also
   tweaked to allow enabling the feature on a host_cpu of amd64 (was only
   allowed on x86_64 before)."

Thanks...

--
    -Chad

Reply via email to