[gpsd-dev] Cleanup patches for the mingw port

[Eric S. Raymond]

Here are the results of my attempts to apply the first round of mingw
patches.

0001-Rename-add_device-to-gpsd_add_device-and-remove-stat.patch
0002-Move-nowait-out-of-gpsd_add_device.patch
0003-Invent-NOWAIT-macro.patch
0004-Add-can_net-to-driver.nmea2000-structure.patch
0005-Add-framework-for-multiple-units.patch

        These look like Reinhardt's work on the CAN support, 
        included on the cleanup branch by mistake.

0006-Ignore-Eclipse-project-files.patch

        Applied.

0007-Remove-ugly-intptr_t-casts-from-gpsdata-gps_fd.patch

        I agree those casts are ugly, but without them one of our
        auditing tools gets confused.  Not applied, but we can discusss
        this further.

0008-Replace-tests-of-value-of-gps_fd-which-thus-assumed-.patch

        Failed to apply cleanly to libgps_core.c.  Please resubmit
        against head.  Actually, what I'd like is a patch that
        does all the GOODSOCK/BADSOCK changes in one go.

0009-Placate-splint-by-adding-U-suffix-to-unsigned-intege.patch

        splint wasn't complaining for me, but I've applied
        a slightly modified version of this patch (u suffixes
        rather than U) because it looks like a good cleanup.

0010-Silence-some-more-splint-warnings-and-use-GOOD-BAD-S.patch

        Partly applied. I omitted the BADSOCK application.
        I wasn't seeing splint warnings from this code, which
        makes me wonder why you were.

0011-Placate-splint-by-conditionally-not-declaring-defini.patch
0012-Replace-more-numeric-comparisons-of-gpsdata-gps_fd-w.patch
0013-More-replacements-of-numeric-comparisons-of-sockets-.patch
0014-More-GOOD-BADSOCK-use.patch
0015-The-last-I-think-of-GOOD-BADSOCK-usage.patch

        Not applied; I'd like to see one patch that rolls up all
        of these, 0008, 0031, and the BADSOCK parts of 0010/0016/0025.

0016-Use-socket_t-rather-than-int-and-GOOD-BADSOCK-rather.patch

        int to socket_t changes applied. 

0017-Remove-use-of-struct-termios-from-public-API-of-seri.patch

        Applied.

0018-sockaddr_t-only-needs-to-accomodate-a-sockaddr_in6-i.patch

        Applied.

0019-Use-UNUSED-rather-than-uselessly-setting-a-parameter.patch

        Applied.

0020-Silence-some-pointer-signedness-warnings.patch

        Applied.

0021-Silence-cppcheck-warning.-Limits-year-values-to-10-9.patch

        Applied.

0022-Silence-splint-and-cppcheck-mostly-by-reducing-varia.patch

        Patch failed. Please resubmit.

0023-Comply-stricly-with-strict-aliasing-rules-by-using-m.patch

        Applied.  Good catch!

0024-Silence-another-cppcheck-warning.patch

        Applied.

0025-Instead-of-passing-errno-via-gps_fd-ensure-netlib_co.patch

        Not applied.  I want to see the parts relating to GOODSOCK/BADSOCK
        split apart from the errno juggling.

0026-Use-strerror-rather-than-a-generic-error-message.-No.patch
0027-Make-libgps_core.c-s-extern-functions-leave-errno-wi.patch
0028-Use-errno-for-diagnostics-rather-than-ab-using-socke.patch
0029-Leave-errno-with-a-meaningful-value-on-return.patch

        Not applied, as these look dependent on 0025 to work.
        I'd like to see a all the errno juggling consolidated
        into one patch with a good explanation of how this changes
        the library's expected behavior.

0030-Use-exit-EXIT_-SUCCESS-FAILURE-rather-than-0-1-respe.patch

        Applied.

0031-More-GOOD-BADSOCK.patch

        Not applied.  I want to see this as part of a consolidated
        patch to do that change everywhere.

0032-Remove-unnecessary-includes-of-termios.h.patch

        Applied.

0033-Replace-uint-with-unsigned-everywhere-obviating-one-.patch

        Not applied.  I'll take an equivalent patch that consistently
        uses 'unsigned int', as opposed to bare 'unsigned'.

0034-Compile-most-code-with-Werror.patch

        Applied.  I am quite pleased thgat this didn't break the build.

0035-Use-unsigned-short-instead-of-ushort-obviating-a-typ.patch

        Applied.

0036-Remove-trailing-whitespace-from-C-and-C-source.patch

        Partly applied; there were a few failures. Feel free
        to submit a supplementary patch.

0037-Silence-splint-warning.-Regression-tests-pass.patch

        Applied.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>

The biggest hypocrites on gun control are those who live in upscale
developments with armed security guards -- and who want to keep other
people from having guns to defend themselves.  But what about
lower-income people living in high-crime, inner city neighborhoods?
Should such people be kept unarmed and helpless, so that limousine
liberals can 'make a statement' by adding to the thousands of gun laws
already on the books?"
        --Thomas Sowell