Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932327AbbEUQL1 (ORCPT ); Thu, 21 May 2015 12:11:27 -0400 Received: from mailapp01.imgtec.com ([195.59.15.196]:15182 "EHLO mailapp01.imgtec.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932187AbbEUQLY convert rfc822-to-8bit (ORCPT ); Thu, 21 May 2015 12:11:24 -0400 From: Leonid Yegoshin To: Paul Burton CC: "linux-mips@linux-mips.org" , "rusty@rustcorp.com.au" , "alexinbeijing@gmail.com" , "david.daney@cavium.com" , "alex@alex-smith.me.uk" , "linux-kernel@vger.kernel.org" , "ralf@linux-mips.org" , James Hogan , Markos Chandras , "macro@linux-mips.org" , "eunb.song@samsung.com" , "manuel.lauss@gmail.com" , "andreas.herrmann@caviumnetworks.com" Subject: Re: [PATCH 1/2] MIPS: MSA: bugfix - disable MSA during thread switch correctly Thread-Topic: [PATCH 1/2] MIPS: MSA: bugfix - disable MSA during thread switch correctly Thread-Index: AQHQkni04te1hTLFYUCPQNg+r9/NYp2GnM8A////ymA= Date: Thu, 21 May 2015 16:11:18 +0000 Message-ID: References: <20150519211222.35859.52798.stgit@ubuntu-yegoshin> <20150519211351.35859.80332.stgit@ubuntu-yegoshin>,<20150521090741.GG13811@NP-P-BURTON> In-Reply-To: <20150521090741.GG13811@NP-P-BURTON> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8BIT MIME-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2775 Lines: 75 Yes, I have a program but it is binary only. If you want to understand why disable_DMA() after resume() doesn't work, search for restoring RA register in resume() after changing SP. - Leonid. Paul Burton wrote: On Tue, May 19, 2015 at 02:13:51PM -0700, Leonid Yegoshin wrote: > During thread cloning the new (child) thread should have MSA disabled even > at first thread entry. So, the code to disable MSA is moved from macro > 'switch_to' to assembler function 'resume' before it switches kernel stack > to 'next' (new) thread. Call of 'disable_msa' after 'resume' in 'switch_to' > macro never called a first time entry into thread. Hi Leonid, I'm not sure I understand what you're trying to say here. Do you have an example of a program that demonstrates the behaviour you believe to be broken? Thanks, Paul > Signed-off-by: Leonid Yegoshin > --- > arch/mips/include/asm/switch_to.h | 1 - > arch/mips/kernel/r4k_switch.S | 6 ++++++ > 2 files changed, 6 insertions(+), 1 deletion(-) > > diff --git a/arch/mips/include/asm/switch_to.h b/arch/mips/include/asm/switch_to.h > index e92d6c4b5ed1..0d0f7f8f8b3a 100644 > --- a/arch/mips/include/asm/switch_to.h > +++ b/arch/mips/include/asm/switch_to.h > @@ -104,7 +104,6 @@ do { \ > if (test_and_clear_tsk_thread_flag(prev, TIF_USEDMSA)) \ > __fpsave = FP_SAVE_VECTOR; \ > (last) = resume(prev, next, task_thread_info(next), __fpsave); \ > - disable_msa(); \ > } while (0) > > #define finish_arch_switch(prev) \ > diff --git a/arch/mips/kernel/r4k_switch.S b/arch/mips/kernel/r4k_switch.S > index 04cbbde3521b..7dbb64656bfe 100644 > --- a/arch/mips/kernel/r4k_switch.S > +++ b/arch/mips/kernel/r4k_switch.S > @@ -25,6 +25,7 @@ > /* preprocessor replaces the fp in ".set fp=64" with $30 otherwise */ > #undef fp > > +#define t4 $12 > /* > * Offset to the current process status flags, the first 32 bytes of the > * stack are not used. > @@ -73,6 +74,11 @@ > cfc1 t1, fcr31 > msa_save_all a0 > .set pop /* SET_HARDFLOAT */ > + li t4, MIPS_CONF5_MSAEN > + mfc0 t3, CP0_CONFIG, 5 > + or t3, t3, t4 > + xor t3, t3, t4 > + mtc0 t3, CP0_CONFIG, 5 > > sw t1, THREAD_FCR31(a0) > b 2f > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/