From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751751Ab0LVBsu (ORCPT ); Tue, 21 Dec 2010 20:48:50 -0500 Received: from ozlabs.org ([203.10.76.45]:44267 "EHLO ozlabs.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751364Ab0LVBst (ORCPT ); Tue, 21 Dec 2010 20:48:49 -0500 From: Rusty Russell To: Dmitry Torokhov Subject: Re: [PATCH] Show version information for built-in modules in sysfs Date: Wed, 22 Dec 2010 12:18:45 +1030 User-Agent: KMail/1.13.5 (Linux/2.6.35-23-generic; KDE/4.5.1; i686; ; ) Cc: LKML References: <20101215220019.GA6560@dtor-ws.eng.vmware.com> <201012221132.41057.rusty@rustcorp.com.au> <20101222011712.GA19105@dtor-ws.eng.vmware.com> In-Reply-To: <20101222011712.GA19105@dtor-ws.eng.vmware.com> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <201012221218.45533.rusty@rustcorp.com.au> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 22 Dec 2010 11:47:12 am Dmitry Torokhov wrote: > On Tue, Dec 21, 2010 at 05:02:40PM -0800, Rusty Russell wrote: > > On Thu, 16 Dec 2010 08:30:19 am Dmitry Torokhov wrote: > > > Currently only drivers that are built as modules have their versions > > > shown in /sys/module//version, but this information might > > > also be useful for built-in drivers as well. This especially important > > > for drivers that do not define any parameters - such drivers, if > > > built-in, are completely invisible from userspace. > > > > > > This patch changes MODULE_VERSION() macro so that in case when we are > > > compiling built-in module, version information is stored in a separate > > > section. Kernel then uses this data to create 'version' sysfs attribute > > > in the same fashion it creates attributes for module parameters. > > > > > > Signed-off-by: Dmitry Torokhov > > > --- > > > > > > This change is driven by our desire to detect whether vmw_balloon driver > > > is built-in into the kernel when we installing our tool package in the > > > guest and avoid installing our own version of the driver. > > > > > > Since vmw_balloon does not have any module parameter nor registers any > > > driver core devices, without this patch it is completely invisible from > > > userspace when built into the kernel. > > > > > > Thanks, > > > Dmitry > > > > > > include/asm-generic/vmlinux.lds.h | 7 ++++ > > > include/linux/module.h | 27 +++++++++++++++ > > > kernel/params.c | 65 ++++++++++++++++++++++++++++++------ > > > 3 files changed, 88 insertions(+), 11 deletions(-) > > > > > > diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h > > > index bd69d79..0d83dd1 100644 > > > --- a/include/asm-generic/vmlinux.lds.h > > > +++ b/include/asm-generic/vmlinux.lds.h > > > @@ -355,6 +355,13 @@ > > > VMLINUX_SYMBOL(__start___param) = .; \ > > > *(__param) \ > > > VMLINUX_SYMBOL(__stop___param) = .; \ > > > + } \ > > > + \ > > > + /* Built-in module versions. */ \ > > > + __modver : AT(ADDR(__modver) - LOAD_OFFSET) { \ > > > + VMLINUX_SYMBOL(__start___modver) = .; \ > > > + *(__modver) \ > > > + VMLINUX_SYMBOL(__stop___modver) = .; \ > > > . = ALIGN((align)); \ > > > VMLINUX_SYMBOL(__end_rodata) = .; \ > > > } \ > > > diff --git a/include/linux/module.h b/include/linux/module.h > > > index 7575bbb..f74ddda 100644 > > > --- a/include/linux/module.h > > > +++ b/include/linux/module.h > > > @@ -58,6 +58,12 @@ struct module_attribute { > > > void (*free)(struct module *); > > > }; > > > > > > +struct module_version_attribute { > > > + struct module_attribute mattr; > > > + const char *module_name; > > > + const char *version; > > > +}; > > > + > > > struct module_kobject > > > { > > > struct kobject kobj; > > > @@ -161,7 +167,28 @@ extern struct module __this_module; > > > Using this automatically adds a checksum of the .c files and the > > > local headers in "srcversion". > > > */ > > > + > > > +#ifdef MODULE > > > #define MODULE_VERSION(_version) MODULE_INFO(version, _version) > > > +#else > > > +#define MODULE_VERSION(_version) \ > > > + extern ssize_t __modver_version_show(struct module_attribute *, \ > > > + struct module *, char *); \ > > > + static struct module_version_attribute __modver_version_attr \ > > > + __used \ > > > + __attribute__ ((unused,__section__ ("__modver"),aligned(sizeof(void *)))) \ > > > > __used and unused seems overkill, and confused. > > > > Must admit that I copied used/unused verbatim from linux/moduleparam.h: > > /* This is the fundamental function for registering boot/module > parameters. */ > #define __module_param_call(prefix, name, ops, arg, isbool, perm) \ > /* Default value instead of permissions? */ \ > static int __param_perm_check_##name __attribute__((unused)) = \ > BUILD_BUG_ON_ZERO((perm) < 0 || (perm) > 0777 || ((perm) & 2)) \ > + BUILD_BUG_ON_ZERO(sizeof(""prefix) > MAX_PARAM_PREFIX_LEN); \ > static const char __param_str_##name[] = prefix #name; \ > static struct kernel_param __moduleparam_const __param_##name \ > __used \ > __attribute__ ((unused,__section__ ("__param"),aligned(sizeof(void *)))) \ > = { __param_str_##name, ops, perm, isbool ? KPARAM_ISBOOL : 0, \ > { arg } } > > > So should it be removed from here as well? I think so, but (as always!) check git blame. Patch welcome! Rusty.