From: Alexander Aring <alex.aring@gmail.com>
To: ptxdist@pengutronix.de
Subject: Re: [ptxdist] [PATCH] libusb: drop libusb and use libusbx
Date: Fri, 28 Mar 2014 14:10:35 +0100 [thread overview]
Message-ID: <20140328131030.GA29424@omega.Speedport_W_921V_1_24_000> (raw)
In-Reply-To: <20140328125539.GX10959@pengutronix.de>
Hi Michael,
On Fri, Mar 28, 2014 at 01:55:39PM +0100, Michael Olbrich wrote:
> On Fri, Mar 28, 2014 at 12:30:29PM +0100, Alexander Aring wrote:
> > This patch drops the libusb support and add the libusbx instead.
> >
> > We can do that because libusb and libusbx offers the same API, original
> > text from libusbx project website:
> >
> > "If you are an existing user of libusb, you don't have to change
> > anything. Just download the libusbx library and use it in your code. The
> > library is designed as a drop in replacement for libusb, so please give
> > it a try. If you don't like it, which we very much doubt, you can always
> > revert..."
> >
> > Another point is that libusbx website says that the libusb project is
> > somehow died and will not longer release new releases:
> >
> > "The reason for the fork is that, despite having dedicated members,
> > libusb has still not been able to produce a new release for the past 2
> > years. When a project fails to produce regular releases, we consider
> > that you, its user, are paying the ultimate price. This is because it
> > means that patches and new feature are being witheld and you end up
> > wasting your time."
> >
> > This patch fix also a few Kconfig bugs which I detected:
> >
> > i) switch to "--disable-static", otherwise it was "--enable-shared
> > --enable-static"
> >
> > ii) correct use of --disable-log and --enable-debug-log, there was an
> > option for these two configure flags but also was a static
> > "--disable-log --disable-debug-log" in the configure call, which
> > makes no sense.
> >
> > iii) also adding some new disable flags like --disable-timerfd,
> > --disable-system-log and --enable-tests-build
>
> Using libusbx sounds good. Any reason for --disable-timerfd and
> --enable-tests-build? I've not tested it yet, but it looks good so far.
>
for disable-timerfd:
I don't know what a timerfd is, fd maybe stands for filedescriptor... I
googled it and it's this:
http://man7.org/linux/man-pages/man2/timerfd_create.2.html
Yea, I can enable it or we don't make anything and let's do configure
the work. timerfd is [default=auto]. Maybe there exists also some LIBC
dependencies like in uclibc, then I need to check that, then we should
do nothing with this flag, or?
for enable-test-builds:
configure says "build example applications [default=no]", I don't know
what for examples it builds but then we need to install these
applications in targetinstall aswell. I think we don't need these examples.
> Michael
>
> > iv) I add "--disable-udev" there but we should have a --enable-udev...
> > The old one has no kind of udev support and I got a recursive
> > dependency detection because udev-legacy needs usbutils which needs
...
> > +20100101 rsc: added as http://libusb.org/ticket/24
> > +
> > +Signed-off-by: Alexander Aring <alex.aring@gmail.com>
> > +---
> > + libusb-1.0.pc.in | 2 +-
> > + 1 file changed, 1 insertion(+), 1 deletion(-)
> > +
> > +diff --git a/libusb-1.0.pc.in b/libusb-1.0.pc.in
> > +index 7bc33c1..ec361a5 100644
> > +--- a/libusb-1.0.pc.in
> > ++++ b/libusb-1.0.pc.in
> > +@@ -6,6 +6,6 @@ includedir=@includedir@
> > + Name: libusbx-1.0
> > + Description: C API for USB device access from Linux, Mac OS X, Windows and OpenBSD/NetBSD userspace
> > + Version: @VERSION@
> > +-Libs: -L${libdir} -lusb-1.0
> > ++Libs: -L${libdir} -lusb-1.0 -lusb-1.0 -lrt -lpthread
yea, I will fix that in v2.
> > + Libs.private: @LIBS@
> > + Cflags: -I${includedir}/libusb-1.0
> > +--
> > +1.9.1
> > +
> > diff --git a/patches/libusbx-1.0.18/series b/patches/libusbx-1.0.18/series
> > new file mode 100644
> > index 0000000..4284244
> > --- /dev/null
> > +++ b/patches/libusbx-1.0.18/series
> > @@ -0,0 +1 @@
...
> > - --disable-examples-build \
> > - --enable-static \
> > + --disable-static \
> > --enable-shared \
> > - --$(call ptx/endis, PTXCONF_LIBUSB_DISABLE_LOG)-log
> > -
> > -ifdef PTXCONF_LIBUSB_DEBUG_LOG
> > -LIBUSB_AUTOCONF += --enable-debug-log
> > -endif
> > + --disable-udev \
> > + --disable-timerfd \
> > + --$(call ptx/disen, PTXCONF_LIBUSB_DISABLE_LOG)-log \
I also change the negative logic here, it was endis before. Maybe we
should change that in PTXCONF_LIBUSB_LOG and there is also one
dependency like PTXCONF_LIBUSB_DEBUG_LOG, but then we can also add a
PTXCONF_LIBUSB_SYSTEM_LOG for the --enable-system-log parameter.
- Alex
--
ptxdist mailing list
ptxdist@pengutronix.de
next prev parent reply other threads:[~2014-03-28 13:10 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-03-28 11:30 Alexander Aring
2014-03-28 11:37 ` Alexander Aring
2014-03-28 12:55 ` Michael Olbrich
2014-03-28 13:10 ` Alexander Aring [this message]
2014-03-28 20:51 ` Alexander Aring
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140328131030.GA29424@omega.Speedport_W_921V_1_24_000 \
--to=alex.aring@gmail.com \
--cc=ptxdist@pengutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox