[Bug 566171] Review Request: libhid - A userspace USB HID acess library

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


https://bugzilla.redhat.com/show_bug.cgi?id=566171

Michael Schwendt <mschwendt at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
         AssignedTo|nobody at fedoraproject.org    |mschwendt at gmail.com
               Flag|                            |fedora-review?

--- Comment #12 from Michael Schwendt <mschwendt at gmail.com> 2010-03-06 08:06:03 EST ---
Re: comment 9

> db2x_docbook2man

Without the patch, it creates a bad man page here. It executes:

xsltproc -''-nonet
/usr/share/sgml/docbook/xsl-stylesheets-1.75.2/manpages/docbook.xsl
libhid-detach-device.dbk

which _is_ a valid path to the stylesheet file:

$ rpm -qf /usr/share/sgml/docbook/xsl-stylesheets-1.75.2/manpages/docbook.xsl
docbook-style-xsl-1.75.2-4.fc12.noarch

But it prints many warnings/errors, and the created manual contains mistakes
and is lacking substitutions.

So, your patch to use db2x_docbook2man creates a good man page.


> libhid-0.2.17-3.fc12.src.rpm

%patch0 must be applied before running autoreconf, or else the generated
Makefile.in would be based on the unpatched Makefile.am


> -make CFLAGS="$CFLAGS" 
> +make %{?_smp_mflags}

Good. This one reveals that ./m4/md_conf_debugging.m4 messes with the compiler
flags. It strips -g from the $CFLAGS by default. And one cannot simply
configure with --enable-debug, because then it modifies the optimisation flags.

The ugly work-around would be to put back the CFLAGS definition when running
"make". The better fix would be to patch the configure script and stop it from
messing with the flags like this.


> -%ifnarch i686
> +# Platforms in which libhid must build fine
> +%ifarch x86_64 ppc64 ppc sparc64

The comment doesn't match the purpose of that %ifarch, though. This %ifarch is
only for those platforms where both macros expand to a separate directory, to
avoid rpmbuild's warning about "duplicate files entries" in the spec file
(comment 5). On "ppc", %{python_sitearch} expands to /usr/lib just like
%{python_sitelib}. Hence you get that warning for a ppc build.


Don't see any other issues with the package.

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.