All of lore.kernel.org
 help / color / mirror / Atom feed
From: Arnout Vandecappelle <arnout@mind.be>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH] lcdapi: new package
Date: Wed, 17 Oct 2012 22:12:15 +0200	[thread overview]
Message-ID: <507F111F.1070107@mind.be> (raw)
In-Reply-To: <1350482898-5980-1-git-send-email-spdawson@gmail.com>

On 17/10/12 16:08, spdawson at gmail.com wrote:
> From: Simon Dawson<spdawson@gmail.com>
>
> Signed-off-by: Simon Dawson<spdawson@gmail.com>
[snip]
> diff --git a/package/lcdapi/lcdapi-make-toString-static.patch b/package/lcdapi/lcdapi-make-toString-static.patch
> new file mode 100644
> index 0000000..0ef95fc
> --- /dev/null
> +++ b/package/lcdapi/lcdapi-make-toString-static.patch
> @@ -0,0 +1,19 @@
> +make the toString function defined in the LCDCallback.h header into a static
> +function, to fix link-time errors like the following.
> +
> +  multiple definition of `toString(char)'
> +
> +Signed-off-by: Simon Dawson<spdawson@gmail.com>
> +
> +diff -Nurp a/keys/LCDCallback.h b/keys/LCDCallback.h
> +--- a/keys/LCDCallback.h	2004-07-27 19:06:27.000000000 +0100
> ++++ b/keys/LCDCallback.h	2012-10-01 10:59:52.685708010 +0100
> +@@ -77,7 +77,7 @@ class LCDCallback
> +
> + typedef std::map<KeyEvent, LCDCallback *>  CallbackMap;
> +
> +-std::string toString(KeyEvent t)
> ++static std::string toString(KeyEvent t)

  Making it inline sounds more appropriate.

  BTW, did you upstream the patches?

> + {
> +   std::string s(1, (char)t);
> +   return s;
> diff --git a/package/lcdapi/lcdapi.mk b/package/lcdapi/lcdapi.mk
> new file mode 100644
> index 0000000..37e276c
> --- /dev/null
> +++ b/package/lcdapi/lcdapi.mk
> @@ -0,0 +1,73 @@
> +#############################################################
> +#
> +# lcdapi
> +#
> +#############################################################
> +LCDAPI_VERSION = 0.2
> +LCDAPI_SITE = ftp://ftp2.c-sait.net/csait
> +LCDAPI_DEPENDENCIES = host-lcdapi

  Why is this needed?  Or rather: I can build without this.
If it is not needed, the whole host-lcdapi can be removed BTW.

> +LCDAPI_INSTALL_STAGING = YES
> +
> +LCDAPI_LICENSE = LGPLv2.1+
> +LCDAPI_LICENSE_FILES = COPYING
> +
> +LCDAPI_MAKE_OPT = \
> +	$(TARGET_CONFIGURE_OPTS) \
> +	LD="$(TARGET_CXX)" \
> +	LDFLAGS="$(TARGET_LDFLAGS) -shared"

  It would make sense to also add CC="$(TARGET_CXX)" (although not strictly
necessary, because gcc recognizes the .cpp extension).

  More importantly, the CFLAGS overrides the CFLAGS of lcdapi's
Makefile, and that contains the all-important -fPIC.  I wonder
how you got it built without that...  So either:

- Set CFLAGS to "$(TARGET_CFLAGS) -fPIC"
- Patch the Makefile to append to CFLAGS
- Set CC to "$(TARGET_CXX) $(TARGET_CFLAGS)" and leave CFLAGS alone

  Finally, it would make sense to call it LCDAPI_MAKE_OPTS,
similar to TARGET_CONFIGURE_OPTS.

> +
> +HOST_LCDAPI_MAKE_OPT = \
> +	$(HOST_CONFIGURE_OPTS) \
> +	LD="$(HOSTCXX)" \
> +	LDFLAGS="$(HOST_LDFLAGS) -shared"
> +
> +define LCDAPI_BUILD_CMDS
> +	$(TARGET_MAKE_ENV) $(MAKE) $(LCDAPI_MAKE_OPT) -C $(@D)
> +endef
> +
> +define HOST_LCDAPI_BUILD_CMDS
> +	$(HOST_MAKE_ENV) $(MAKE) $(HOST_LCDAPI_MAKE_OPT) -C $(@D)
> +endef
> +
> +define DO_LCDAPI_INSTALL
> +	$(INSTALL) -m 0755 -d $(1)/usr/lib
> +	$(INSTALL) -m 0755 -t $(1)/usr/lib $(@D)/lib/liblcdapi.so

  We would typically use
	$(INSTALL) -m 0755 -D $(@D)/lib/liblcdapi.so $(1)/usr/lib
(which does the mkdir and the install in one shot).

> +	$(INSTALL) -m 0755 -d $(1)/usr/include/lcdapi
> +	for i in api include keys sensors; do \
> +		$(INSTALL) -m 0755 -d $(1)/usr/include/lcdapi/$$i&&  \
> +		$(INSTALL) -m 0644 -t $(1)/usr/include/lcdapi/$$i \
> +			$(@D)/$$i/*.h; \

  Same here, but then iterating over api/*.h include/*.h keys/*.h sensors/*.h


  But wouldn't it be cleaner to add (and upstream) an 'install' target
to the Makefile?


  Regards,
  Arnout

> +	done
> +endef
> +
> +define DO_LCDAPI_UNINSTALL
> +	$(RM) $(1)/usr/lib/liblcdapi.so
> +	$(RM) -r $(1)/usr/include/lcdapi
> +endef
> +
> +define HOST_LCDAPI_INSTALL_CMDS
> +	$(call DO_LCDAPI_INSTALL,$(HOST_DIR))
> +endef
> +
> +define LCDAPI_INSTALL_STAGING_CMDS
> +	$(call DO_LCDAPI_INSTALL,$(STAGING_DIR))
> +endef
> +
> +define LCDAPI_INSTALL_TARGET_CMDS
> +	$(call DO_LCDAPI_INSTALL,$(TARGET_DIR))
> +endef
> +
> +define LCDAPI_UNINSTALL_STAGING_CMDS
> +	$(call DO_LCDAPI_UNINSTALL,$(STAGING_DIR))
> +endef
> +
> +define LCDAPI_UNINSTALL_TARGET_CMDS
> +	$(call DO_LCDAPI_UNINSTALL,$(TARGET_DIR))
> +endef
> +
> +define LCDAPI_CLEAN_CMDS
> +	$(TARGET_MAKE_ENV) $(MAKE) $(LCDAPI_MAKE_OPT) -C $(@D) clean
> +endef
> +
> +$(eval $(generic-package))
> +$(eval $(host-generic-package))

-- 
Arnout Vandecappelle                               arnout at mind be
Senior Embedded Software Architect                 +32-16-286540
Essensium/Mind                                     http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium                BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F

  reply	other threads:[~2012-10-17 20:12 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-10-17 14:08 [Buildroot] [PATCH] lcdapi: new package spdawson at gmail.com
2012-10-17 20:12 ` Arnout Vandecappelle [this message]
2012-10-18  9:39   ` Simon Dawson
2012-10-19 20:04     ` Arnout Vandecappelle
2012-10-24  9:07       ` Simon Dawson

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=507F111F.1070107@mind.be \
    --to=arnout@mind.be \
    --cc=buildroot@busybox.net \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.