On Wed, Jun 20, 2012 at 7:33 AM, Pavel Shilovsky <pshilovsky@xxxxxxxxx> wrote: > 2012/6/20 Pavel Shilovsky <pshilovsky@xxxxxxxxx>: >> 2012/6/20 Jeff Layton <jlayton@xxxxxxxxx>: >>> On Tue, 19 Jun 2012 16:39:50 +0400 >>> Pavel Shilovsky <pshilovsky@xxxxxxxxx> wrote: >>> >>>> +#define STATUS_GUARD_PAGE_VIOLATION 0x80000001 >>>> +#define STATUS_DATATYPE_MISALIGNMENT 0x80000002 >>>> +#define STATUS_BREAKPOINT 0x80000003 >>>> +#define STATUS_SINGLE_STEP 0x80000004 >>>> +/* #define STATUS_BUFFER_OVERFLOW 0x80000005 */ >>>> +#define STATUS_NO_MORE_FILES 0x80000006 >>>> +#define STATUS_WAKE_SYSTEM_DEBUGGER 0x80000007 >>>> +#define STATUS_HANDLES_CLOSED 0x8000000A >>>> +#define STATUS_NO_INHERITANCE 0x8000000B >>>> +#define STATUS_GUID_SUBSTITUTION_MADE 0x8000000C >>> >>> Well, on the plus side, you did comment out STATUS_BUFFER_OVERFLOW >>> which was a definite conflict with the definitions in the SMB1 status >>> code list. However...I'm fairly certain that there are other conflicts >>> here, so that was just an example. >>> >>> You really have two choices here: >>> >>> 1) make smb2status.h and nterr.h be mutually exclusive. Ensure that you >>> never #include both files. At first glance, this may be easy, but the >>> trick will be to ensure that no one does this in the future. >>> >>> 2) rename all of these error codes such that they do not conflict with >>> the SMB1 errors. This is more certain to ensure that you don't have >>> problems, but it's a bigger change. >>> >>> I really don't care which one you choose, but we really do need to take >>> steps to ensure that we do not get conflicts here. >>> >>> -- >>> Jeff Layton <jlayton@xxxxxxxxx> >> >> nterr.h has only 5 macros started with STATUS_: >> >> #define STATUS_MORE_ENTRIES 0x0105 >> #define STATUS_1804>---- 0x070c >> #define STATUS_NOTIFY_ENUM_DIR 0x010c >> #define STATUS_SOME_UNMAPPED 0x0107 >> #define STATUS_BUFFER_OVERFLOW 0x80000005 >> >> I've already checked all them and commented dublicates in smb2status.h >> (STATUS_MORE_ENTRIES, STATUS_NOTIFY_ENUM_DIR, STATUS_BUFFER_OVERFLOW) >> - no conflicts now. >> >> But I don't care to rename all error codes to something like >> SMB2_STATUS_ to make sure we will not miss something in future. >> >> -- >> Best regards, >> Pavel Shilovsky. > > After a discussion with Jeff we came to the following rules: > 1) prefix 7 error codes from nterr.h with NT_ that doesn't thave NT_ > and consider them as CIFS codes. > 2) consider codes from smb2status.h patch as new style codes (suitable > for SMB2, SMB2.1 and SMB3). Yes - but also have to doublecheck that we didn't miss some new ones. -- Thanks, Steve -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
