This patch fixes a sparse warning on layout.c (ptlrpc) that was caused by having preprocessor directives in the arguments to a macro.
Signed-off-by: Filipe Gonçalves <[email protected]>
---
drivers/staging/lustre/lustre/ptlrpc/layout.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/staging/lustre/lustre/ptlrpc/layout.c b/drivers/staging/lustre/lustre/ptlrpc/layout.c
index 5b83371..211df78 100644
--- a/drivers/staging/lustre/lustre/ptlrpc/layout.c
+++ b/drivers/staging/lustre/lustre/ptlrpc/layout.c
@@ -978,10 +978,11 @@ struct req_msg_field RMF_CONN =
EXPORT_SYMBOL(RMF_CONN);
struct req_msg_field RMF_CONNECT_DATA =
+#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
DEFINE_MSGF("cdata",
RMF_F_NO_SIZE_CHECK /* we allow extra space for interop */,
-#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
sizeof(struct obd_connect_data),
+ lustre_swab_connect, NULL);
#else
/* For interoperability with 1.8 and 2.0 clients/servers.
* The RPC verification code allows larger RPC buffers, but not
@@ -990,9 +991,11 @@ struct req_msg_field RMF_CONNECT_DATA =
* size is at least as large as obd_connect_data_v1. That is not
* not in itself harmful, since the chance of just corrupting this
* field is low. See JIRA LU-16 for details. */
+ DEFINE_MSGF("cdata",
+ RMF_F_NO_SIZE_CHECK /* we allow extra space for interop */,
sizeof(struct obd_connect_data_v1),
-#endif
lustre_swab_connect, NULL);
+#endif
EXPORT_SYMBOL(RMF_CONNECT_DATA);
struct req_msg_field RMF_DLM_REQ =
--
1.9.1
On Sat, Oct 11, 2014 at 08:13:42PM +0100, Filipe Gon?alves wrote:
> This patch fixes a sparse warning on layout.c (ptlrpc) that was caused by having preprocessor directives in the arguments to a macro.
>
> Signed-off-by: Filipe Gon?alves <[email protected]>
> ---
> drivers/staging/lustre/lustre/ptlrpc/layout.c | 7 +++++--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/staging/lustre/lustre/ptlrpc/layout.c b/drivers/staging/lustre/lustre/ptlrpc/layout.c
> index 5b83371..211df78 100644
> --- a/drivers/staging/lustre/lustre/ptlrpc/layout.c
> +++ b/drivers/staging/lustre/lustre/ptlrpc/layout.c
> @@ -978,10 +978,11 @@ struct req_msg_field RMF_CONN =
> EXPORT_SYMBOL(RMF_CONN);
>
> struct req_msg_field RMF_CONNECT_DATA =
> +#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
> DEFINE_MSGF("cdata",
> RMF_F_NO_SIZE_CHECK /* we allow extra space for interop */,
> -#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
> sizeof(struct obd_connect_data),
> + lustre_swab_connect, NULL);
Ick ick ick.
Yeah, sparse might complain about this, but how about just properly
deleting the #ifdef entirely, and not perpetuate it even more?
It shouldn't be needed anymore now that the code is in the kernel tree.
greg k-h
Hello!
No, it's not the way to test the kernel version, it's the way to test internal
lustre version.
Either way maintaining compatibility with Lustre 1.8 and 2.0 servers should not
be important anymore, so it's fine to drop this check indeed.
Bye,
Oleg
On Oct 11, 2014, at 5:06 PM, Filipe Gon?alves wrote:
> Ah .. right! I didn't know what OBD_OCD_VERSION() was. Now I see it's
> a way to test kernel version. I am going to submit a new patch
> shortly.
>
> Thanks,
> Filipe
>
> On Sat, Oct 11, 2014 at 9:15 PM, Greg KH <[email protected]> wrote:
>>
>> On Sat, Oct 11, 2014 at 08:13:42PM +0100, Filipe Gon?alves wrote:
>>> This patch fixes a sparse warning on layout.c (ptlrpc) that was caused by having preprocessor directives in the arguments to a macro.
>>>
>>> Signed-off-by: Filipe Gon?alves <[email protected]>
>>> ---
>>> drivers/staging/lustre/lustre/ptlrpc/layout.c | 7 +++++--
>>> 1 file changed, 5 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/staging/lustre/lustre/ptlrpc/layout.c b/drivers/staging/lustre/lustre/ptlrpc/layout.c
>>> index 5b83371..211df78 100644
>>> --- a/drivers/staging/lustre/lustre/ptlrpc/layout.c
>>> +++ b/drivers/staging/lustre/lustre/ptlrpc/layout.c
>>> @@ -978,10 +978,11 @@ struct req_msg_field RMF_CONN =
>>> EXPORT_SYMBOL(RMF_CONN);
>>>
>>> struct req_msg_field RMF_CONNECT_DATA =
>>> +#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
>>> DEFINE_MSGF("cdata",
>>> RMF_F_NO_SIZE_CHECK /* we allow extra space for interop */,
>>> -#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
>>> sizeof(struct obd_connect_data),
>>> + lustre_swab_connect, NULL);
>>
>> Ick ick ick.
>>
>> Yeah, sparse might complain about this, but how about just properly
>> deleting the #ifdef entirely, and not perpetuate it even more?
>>
>> It shouldn't be needed anymore now that the code is in the kernel tree.
>>
>> greg k-h
Hi,
Oops.. sorry for the mess then. I am still trying to get to know the
codebase. I submitted a new patch where I removed the check.
Thanks for your patience :)
Best,
Filipe
On Sat, Oct 11, 2014 at 10:49 PM, Drokin, Oleg <[email protected]> wrote:
>
> Hello!
>
> No, it's not the way to test the kernel version, it's the way to test internal
> lustre version.
> Either way maintaining compatibility with Lustre 1.8 and 2.0 servers should not
> be important anymore, so it's fine to drop this check indeed.
>
> Bye,
> Oleg
> On Oct 11, 2014, at 5:06 PM, Filipe Gonçalves wrote:
>
>> Ah .. right! I didn't know what OBD_OCD_VERSION() was. Now I see it's
>> a way to test kernel version. I am going to submit a new patch
>> shortly.
>>
>> Thanks,
>> Filipe
>>
>> On Sat, Oct 11, 2014 at 9:15 PM, Greg KH <[email protected]> wrote:
>>>
>>> On Sat, Oct 11, 2014 at 08:13:42PM +0100, Filipe Gonçalves wrote:
>>>> This patch fixes a sparse warning on layout.c (ptlrpc) that was caused by having preprocessor directives in the arguments to a macro.
>>>>
>>>> Signed-off-by: Filipe Gonçalves <[email protected]>
>>>> ---
>>>> drivers/staging/lustre/lustre/ptlrpc/layout.c | 7 +++++--
>>>> 1 file changed, 5 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/drivers/staging/lustre/lustre/ptlrpc/layout.c b/drivers/staging/lustre/lustre/ptlrpc/layout.c
>>>> index 5b83371..211df78 100644
>>>> --- a/drivers/staging/lustre/lustre/ptlrpc/layout.c
>>>> +++ b/drivers/staging/lustre/lustre/ptlrpc/layout.c
>>>> @@ -978,10 +978,11 @@ struct req_msg_field RMF_CONN =
>>>> EXPORT_SYMBOL(RMF_CONN);
>>>>
>>>> struct req_msg_field RMF_CONNECT_DATA =
>>>> +#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
>>>> DEFINE_MSGF("cdata",
>>>> RMF_F_NO_SIZE_CHECK /* we allow extra space for interop */,
>>>> -#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
>>>> sizeof(struct obd_connect_data),
>>>> + lustre_swab_connect, NULL);
>>>
>>> Ick ick ick.
>>>
>>> Yeah, sparse might complain about this, but how about just properly
>>> deleting the #ifdef entirely, and not perpetuate it even more?
>>>
>>> It shouldn't be needed anymore now that the code is in the kernel tree.
>>>
>>> greg k-h
>
Ah .. right! I didn't know what OBD_OCD_VERSION() was. Now I see it's
a way to test kernel version. I am going to submit a new patch
shortly.
Thanks,
Filipe
On Sat, Oct 11, 2014 at 9:15 PM, Greg KH <[email protected]> wrote:
>
> On Sat, Oct 11, 2014 at 08:13:42PM +0100, Filipe Gonçalves wrote:
>> This patch fixes a sparse warning on layout.c (ptlrpc) that was caused by having preprocessor directives in the arguments to a macro.
>>
>> Signed-off-by: Filipe Gonçalves <[email protected]>
>> ---
>> drivers/staging/lustre/lustre/ptlrpc/layout.c | 7 +++++--
>> 1 file changed, 5 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/staging/lustre/lustre/ptlrpc/layout.c b/drivers/staging/lustre/lustre/ptlrpc/layout.c
>> index 5b83371..211df78 100644
>> --- a/drivers/staging/lustre/lustre/ptlrpc/layout.c
>> +++ b/drivers/staging/lustre/lustre/ptlrpc/layout.c
>> @@ -978,10 +978,11 @@ struct req_msg_field RMF_CONN =
>> EXPORT_SYMBOL(RMF_CONN);
>>
>> struct req_msg_field RMF_CONNECT_DATA =
>> +#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
>> DEFINE_MSGF("cdata",
>> RMF_F_NO_SIZE_CHECK /* we allow extra space for interop */,
>> -#if LUSTRE_VERSION_CODE > OBD_OCD_VERSION(2, 7, 50, 0)
>> sizeof(struct obd_connect_data),
>> + lustre_swab_connect, NULL);
>
> Ick ick ick.
>
> Yeah, sparse might complain about this, but how about just properly
> deleting the #ifdef entirely, and not perpetuate it even more?
>
> It shouldn't be needed anymore now that the code is in the kernel tree.
>
> greg k-h