From: Ingo Molnar Date: Sun, 05 Sep 2010 12:32:29 +0000 Subject: Re: [lm-sensors] [PATCH] therm_throt.c: Fix error handling in Message-Id: <20100905123229.GA8254@elte.hu> List-Id: References: <1283474726-20109-1-git-send-email-fenghua.yu@intel.com> <20100903093553.GA6821@elte.hu> <20100903213456.GA22124@fenghua-desk.sc.intel.com> In-Reply-To: <20100903213456.GA22124@fenghua-desk.sc.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Fenghua Yu Cc: Thomas Gleixner , H Peter Anvin , Len Brown , Jin Dongming , Hidetoshi Seto , Jean Delvare , linux-kernel , lm-sensors * Fenghua Yu wrote: > Or other options could be: > > 1. Just calling sysfs_add_file_to_group() without collecting returned error and > return 0 at the end (driver/pci/pcie/aspm.c does like this). The drawback is > there is no error logged if an unlikely errorr occurs. But user can see some > files are missing in sysfs. > > 2. Or collect errors in err1, err2, etc for each sysfs_add_file_to_group. At > the end, return -ENODEV(??) if any err1, err2, etc is not 0. This option makes > code unreasonable complex to handle unlikely errors. Well, the usual way to handle errors is to abort the operation when it occurs, and return the error code that sysfs_add_file_to_group() gave. The error is not 'fatal' but missing sysfs files sure are confusing, and might break user-land which depends on them. So we should either initialize a driver fully - or not intialize it at all. Now, a sub-case is the question whether to emit something more than the return code from sysfs_add_file_to_group(). If it's exceedingly rare (and subsequently poorly tested) then adding a WARN_ON_ONCE(ret) is OK - but that error code should be returned. Am i missing any detail? Thanks, Ingo _______________________________________________ lm-sensors mailing list lm-sensors@lm-sensors.org http://lists.lm-sensors.org/mailman/listinfo/lm-sensors