From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from buildserver.ru.mvista.com (unknown [85.21.88.6]) by ozlabs.org (Postfix) with ESMTP id A2F17DDE31 for ; Tue, 23 Jan 2007 01:05:26 +1100 (EST) Date: Mon, 22 Jan 2007 16:43:16 +0300 From: Vitaly Bordug To: Paul Mackerras Subject: Re: [PATCH 1/5] [POWERPC] cpm2: CPM2 interrupt controller fix Message-ID: <20070122164316.79569af9@localhost.localdomain> In-Reply-To: <17844.35471.554647.189851@cargo.ozlabs.ibm.com> References: <20070113004150.1224.58427.stgit@localhost.localdomain> <17844.35471.554647.189851@cargo.ozlabs.ibm.com> Mime-Version: 1.0 Content-Type: multipart/signed; boundary="Sig_edSiHX8P=JF3+RwdHD1.MzJ"; protocol="application/pgp-signature"; micalg=PGP-SHA1 Cc: linuxppc-dev List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , --Sig_edSiHX8P=JF3+RwdHD1.MzJ Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable On Mon, 22 Jan 2007 20:57:35 +1100 Paul Mackerras wrote: > Vitaly Bordug writes: >=20 > > This contains important fixes for the CPM2 PIC code. Eliminated > > CPM_IRQ_OFFSET, pulling the respective interrupt numbers from the > > interrupt mapping. Updated devicetree files to reflect that. > > Changed direct IC-related IO accesses to the IO accessors. Fixed > > all the sense values to keep coherency with ipic. In the current > > code, CPM2 stuff will have no IRQs and hence could be hardly usable. >=20 > You seem to be adding significant stuff; include/asm-powerpc/mpc8260.h > is new, and there is this hunk in your patch 4/5: As I said, new io.h in powerpc broke compilation for both 8272 and 8560 tha= t were committted. >=20 > > diff --git a/arch/powerpc/platforms/Makefile > > b/arch/powerpc/platforms/Makefile index 507d1b9..f750212 100644 > > --- a/arch/powerpc/platforms/Makefile > > +++ b/arch/powerpc/platforms/Makefile > > @@ -8,6 +8,7 @@ endif > > obj-$(CONFIG_PPC_MPC52xx) +=3D 52xx/ > > obj-$(CONFIG_PPC_CHRP) +=3D chrp/ > > obj-$(CONFIG_4xx) +=3D 4xx/ > > +obj-$(CONFIG_PPC_82xx) +=3D 82xx/ yes, as it was missed from the original primer 82xx commit. All the stuff i= s already there so this is nothing but a fix for (a silly) miss.=20 >=20 > Also, this stuff seems to be used on 85xx, and I have not heard from > Kumar or anyone else working on 85xx as to how these changes are going > to affect 85xx. So I'm sorry, but I'm not pushing these changes in > for 2.6.20. >=20 Hmm? I did the IRQ senses generalisation as per Kumar's request - this in f= act is resubmit of the same. > Also, with the changes like this one: >=20 > > - irq_nr -=3D CPM_IRQ_OFFSET; > > + unsigned int irq_nr =3D (unsigned int)irq_map[irq].hwirq; >=20 > you should probably use the virq_to_hw() function defined in > include/asm-powerpc/irq.h, and you should definitely lose the > unnecessary (unsigned int) cast. >=20 OK, I'll try to use virq...() stuff instead. > > +#ifndef _ISA_MEM_BASE > > +#define _ISA_MEM_BASE 0 > > +#endif > > + > > +#ifndef PCI_DRAM_OFFSET > > +#define PCI_DRAM_OFFSET 0 > > +#endif >=20 > Why do we need to have these things defined in mpc8260.h? Shouldn't > we get the correct definition from io.h? >=20 Will check and clean if it is not required. Thanks,=20 -Vitaly --Sig_edSiHX8P=JF3+RwdHD1.MzJ Content-Type: application/pgp-signature; name=signature.asc Content-Disposition: attachment; filename=signature.asc -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFFtL91uOg9JvQhSEsRAnxSAJ9oCyctXJ6TebsIQtniqX2eESBiXgCeK8MD Trr0QaSUx/o8USCPZ2bu5zM= =tNRL -----END PGP SIGNATURE----- --Sig_edSiHX8P=JF3+RwdHD1.MzJ--