From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephan Mueller Subject: Re: #pragma GCC warnings (was: Re: [PATCH] crypto: drbg - use pragmas for disabling optimization) Date: Tue, 23 Jun 2015 11:45:29 +0200 Message-ID: <2009251.BuUdnce5se@tauon.atsec.com> References: <20150623092106.GA18791@gondor.apana.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7Bit Return-path: In-Reply-To: <20150623092106.GA18791@gondor.apana.org.au> Sender: linux-crypto-owner@vger.kernel.org To: Herbert Xu Cc: Geert Uytterhoeven , Guenter Roeck , Peter Zijlstra , Andy Shevchenko , Jim Davis , Stephen Rothwell , Linux-Next , "linux-kernel@vger.kernel.org" , "David S. Miller" , linux-crypto@vger.kernel.org, Waiman Long , Ingo Molnar , x86 List-Id: linux-next.vger.kernel.org Am Dienstag, 23. Juni 2015, 17:21:06 schrieb Herbert Xu: Hi Herbert, >On Tue, Jun 23, 2015 at 11:17:23AM +0200, Geert Uytterhoeven wrote: >> I get that too with m68k-linux-gcc-4.6.3 and m68k-linux-gcc-4.9.0. >> >> With m68k-linux-gnu-gcc-4.1, which is still my default cross-compiler due >> to the good unused warning reporting, I get: >> >> crypto/jitterentropy.c:235: warning: ignoring #pragma GCC push_options >> crypto/jitterentropy.c:236: warning: ignoring #pragma GCC optimize >> crypto/jitterentropy.c:266: warning: ignoring #pragma GCC pop_options >> crypto/jitterentropy.c:295: warning: ignoring #pragma GCC push_options >> crypto/jitterentropy.c:296: warning: ignoring #pragma GCC optimize >> crypto/jitterentropy.c:336: warning: ignoring #pragma GCC pop_options >> crypto/jitterentropy.c:385: warning: ignoring #pragma GCC push_options >> crypto/jitterentropy.c:386: warning: ignoring #pragma GCC optimize >> crypto/jitterentropy.c:416: warning: ignoring #pragma GCC pop_options >> crypto/jitterentropy.c:517: warning: ignoring #pragma GCC push_options >> crypto/jitterentropy.c:518: warning: ignoring #pragma GCC optimize >> crypto/jitterentropy.c:580: warning: ignoring #pragma GCC pop_options > >Stephan, could you look into moving the relevant functions into >its own file which can then be compiled with -O0? Obviously any >dependency on kernel header files would have to be hidden using >functions outside of this file. I will look into it. Thanks Ciao Stephan