From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from sipsolutions.net (crystal.sipsolutions.net [195.210.38.204]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTP id 5779BDDE44 for ; Fri, 6 Jul 2007 20:58:39 +1000 (EST) Subject: Re: powerpc stacktrace and lockdep support From: Johannes Berg To: Sergei Shtylyov In-Reply-To: <468D45CD.1080008@ru.mvista.com> References: <20070108135458.GA23395@lst.de> <1182898073.4769.6.camel@johannes.berg> <1182970829.4769.62.camel@johannes.berg> <1183047642.3735.0.camel@johannes.berg> <468D45CD.1080008@ru.mvista.com> Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="=-Myqp90QvsM4DSznWiTWo" Date: Fri, 06 Jul 2007 12:59:37 +0200 Message-Id: <1183719577.3818.115.camel@johannes.berg> Mime-Version: 1.0 Cc: linuxppc-dev@ozlabs.org, Christoph Hellwig List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , --=-Myqp90QvsM4DSznWiTWo Content-Type: text/plain Content-Transfer-Encoding: quoted-printable Sergei, Thanks for your review! > I suggest to also remove these 3 lines from the heading comment in th= is=20 > file -- they're not true anyway: >=20 > * This file gets included from lowlevel asm headers too, to provide > * wrapped versions of the local_irq_*() APIs, based on the > * raw_local_irq_*() macros from the lowlevel headers. Good point. > > +#ifdef CONFIG_TRACE_IRQFLAGS > > +#define TRACE_DISABLE_INTS bl .powerpc_trace_hardirqs_off > > +#else > > +#define TRACE_DISABLE_INTS > > +#endif > > + >=20 > Erm, weren't those supposed to be in ? I guess they could be there. I was mucking my way through ;) > The following code seemed over-engineered: Yeah, perfectly correct. See my last mail to this thread from this morning. > Again, could have been more compact... unless trace_hardirqs_*() call= s=20 > need to be in certain order WRT writes to PACASOFTIRQEN -- if so, in the = 1st=20 > case that I've pointed out the order was not identical (probably wrong?).= .. Yeah, there is an ordering requirement and it was wrong in that version of the patch. > > +/* > > + * crappy helper for irq-trace > > + */ > > + > > +#include > > +#include > > + > > +#define STACKSPACE GPR0 + 16*8 >=20 > I guess this should be 16*4 for PPC32. > Didn't you forget to add GPR0 to the offsets here? > You certainly did. I bet you're clobbering GPR11/12 (if I didn't misc= ount)... All correct. That might even be why it didn't work on 32-bit for me and then I gave up. In any case, I completely rewrote this file. johannes --=-Myqp90QvsM4DSznWiTWo Content-Type: application/pgp-signature; name=signature.asc Content-Description: This is a digitally signed message part -----BEGIN PGP SIGNATURE----- Comment: Johannes Berg (powerbook) iD8DBQBGjiCZ/ETPhpq3jKURAjJ2AKCjG0p622sEIqsbBz6JjlpqwhAOVwCfQ6sT KquAbapmxljjQniHaiBxpzs= =LKyq -----END PGP SIGNATURE----- --=-Myqp90QvsM4DSznWiTWo--