Jan Kara <jack@xxxxxxx> writes: > On Thu 16-02-12 13:04:44, Alex Elder wrote: >> On Thu, 2012-02-16 at 14:46 +0100, Jan Kara wrote: >> > CC: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> >> > Signed-off-by: Jan Kara <jack@xxxxxxx> >> > --- >> > fs/sysfs/bin.c | 2 ++ >> > 1 files changed, 2 insertions(+), 0 deletions(-) >> > >> > diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c >> > index a475983..6ceb16f 100644 >> > --- a/fs/sysfs/bin.c >> > +++ b/fs/sysfs/bin.c >> > @@ -225,6 +225,8 @@ static int bin_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) >> > if (!sysfs_get_active(attr_sd)) >> > return VM_FAULT_SIGBUS; >> > >> > + file_update_time(file); >> > + >> > ret = 0; >> > if (bb->vm_ops->page_mkwrite) >> > ret = bb->vm_ops->page_mkwrite(vma, vmf); >> >> If the filesystem's page_mkwrite() function is responsible >> for updating the time, can't the call to file_update_time() >> here be conditional? >> >> I.e: >> ret = 0; >> if (bb->vm_ops->page_mkwrite) >> ret = bb->vm_ops->page_mkwrite(vma, vmf); >> else >> file_update_time(file); > Hmm, I didn't look previously where do we get bb->vm_ops. It seems they > are inherited from vma->vm_ops so what you suggest should be safe without > any further changes. So I can do that if someone who understands the sysfs > code likes it more. I do. Essentially sysfs is being a stackable filesystem here, because sysfs needs the ability to remove a file mapping. In practice we could probably get away without a single file_update_time(file) here because there are mmio mappings. Normally for pci resources, but we might as well use good form since we can. Eric -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>