From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp-out9.tpgi.com.au ([220.244.226.119] helo=mail9.tpgi.com.au) by linuxtogo.org with esmtp (Exim 4.72) (envelope-from ) id 1R4NeC-0001ou-Py for openembedded-core@lists.openembedded.org; Fri, 16 Sep 2011 03:55:21 +0200 X-TPG-Junk-Checked: Yes X-TPG-Junk-Status: Message not scanned because user authenticated using SMTP AUTH X-TPG-Antivirus: Passed X-TPG-Abuse: host=220-244-174-13.static.tpgi.com.au; ip=220.244.174.13; date=Fri, 16 Sep 2011 11:28:31 +1000; auth=jN+vu9lIznpdsLAI8D8JiNICYhl1zChPtcoQj/CnROE= Received: from dev1.localnet (220-244-174-13.static.tpgi.com.au [220.244.174.13]) (authenticated bits=0) by mail9.tpgi.com.au (envelope-from marc@cpdesign.com.au) (8.14.3/8.14.3) with ESMTP id p8G1STgB002961 for ; Fri, 16 Sep 2011 11:28:31 +1000 From: Marc Reilly Organization: Creative Product Designs To: "Patches and discussions about the oe-core layer" Date: Fri, 16 Sep 2011 11:31:37 +1000 User-Agent: KMail/1.13.6 (Linux/2.6.37.6-0.7-desktop; KDE/4.6.0; x86_64; ; ) References: <1316078776-19890-1-git-send-email-denis@eukrea.com> <1316078776-19890-3-git-send-email-denis@eukrea.com> In-Reply-To: <1316078776-19890-3-git-send-email-denis@eukrea.com> X-KMail-Markup: true MIME-Version: 1.0 Message-Id: <201109161131.37982.marc@cpdesign.com.au> Subject: Re: [PATCH 2/2] qt4(embedded and x11): disable neon for armv6-vfp X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.11 Precedence: list Reply-To: Patches and discussions about the oe-core layer List-Id: Patches and discussions about the oe-core layer List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 16 Sep 2011 01:55:21 -0000 Content-Type: multipart/alternative; boundary="Boundary-01=_5bqcOMIJHzWnwx9" Content-Transfer-Encoding: 7bit --Boundary-01=_5bqcOMIJHzWnwx9 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit On Thursday, September 15, 2011 07:26:16 PM Denis Carikli wrote: > without the -no-neon flag, neon is "autodetected" > by looking if the compiler is capable of compiling > a neon test, and succeed, and neon is then enabled > during the compilation. > > Signed-off-by: Denis Carikli > QT_CONFIG_FLAGS_append_armv6 = " -no-neon " > +QT_CONFIG_FLAGS_append_armv6-vfp = " -no-neon " I also ran into this problem, with the same fix. There are a couple of other places with armv6 overrides that could misbehave, notably: TARGET_CC_ARCH_append_armv6 = " -D__SOFTFP__" (python_2.6.6.bb) QEMU_OPTIONS_armv6 = "-cpu arm1136" (bitbake.conf) Is there a way to do this more generically?, something like: QT_CONFIG_FLAGS_append_armv6* = " ... " Cheers Marc --Boundary-01=_5bqcOMIJHzWnwx9 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: 7bit

On Thursday, September 15, 2011 07:26:16 PM Denis Carikli wrote:

> without the -no-neon flag, neon is "autodetected"

> by looking if the compiler is capable of compiling

> a neon test, and succeed, and neon is then enabled

> during the compilation.

>

> Signed-off-by: Denis Carikli <denis@eukrea.com>



> QT_CONFIG_FLAGS_append_armv6 = " -no-neon "

> +QT_CONFIG_FLAGS_append_armv6-vfp = " -no-neon "


I also ran into this problem, with the same fix.



There are a couple of other places with armv6 overrides that could misbehave, notably:

TARGET_CC_ARCH_append_armv6 = " -D__SOFTFP__"

(python_2.6.6.bb)

QEMU_OPTIONS_armv6 = "-cpu arm1136"

(bitbake.conf)


Is there a way to do this more generically?, something like:

QT_CONFIG_FLAGS_append_armv6* = " ... "


Cheers

Marc



--Boundary-01=_5bqcOMIJHzWnwx9--