The previous Kconfig patch led to some other build errors as
reported by the 0day bot and my own overnight build testing.
These are all in <linux/skbuff.h> when KCOV is enabled but
SKB_EXTENSIONS is not enabled, so fix those by combining those conditions
in the header file.
Also, add stubs for skb_ext_add() and skb_ext_find() to reduce the
amount of ifdef-ery. (Jakub)
Fixes: 6370cc3bbd8a ("net: add kcov handle to skb extensions")
Fixes: 85ce50d337d1 ("net: kcov: don't select SKB_EXTENSIONS when there is no NET")
Signed-off-by: Randy Dunlap <[email protected]>
Reported-by: kernel test robot <[email protected]>
Cc: Aleksandr Nogikh <[email protected]>
Cc: Willem de Bruijn <[email protected]>
Cc: Jakub Kicinski <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: Matthieu Baerts <[email protected]>
---
v4: The enum for SKB_EXT_KCOV_HANDLE needs to be exposed unconditionally
because it is used in skb_get/set_kcov_handle(), which are always
present since v3.
v3: (as suggested by Jakub Kicinski <[email protected]>)
add stubs for skb_ext_add() and skb_ext_find() to reduce the ifdef-ery
v2: (as suggested by Matthieu Baerts <[email protected]>)
drop an extraneous space in a comment;
use CONFIG_SKB_EXTENSIONS instead of CONFIG_NET;
include/linux/skbuff.h | 14 +++++---------
1 file changed, 5 insertions(+), 9 deletions(-)
--- linux-next-20201113.orig/include/linux/skbuff.h
+++ linux-next-20201113/include/linux/skbuff.h
@@ -4137,7 +4137,6 @@ static inline void skb_set_nfct(struct s
#endif
}
-#ifdef CONFIG_SKB_EXTENSIONS
enum skb_ext_id {
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
SKB_EXT_BRIDGE_NF,
@@ -4151,12 +4150,11 @@ enum skb_ext_id {
#if IS_ENABLED(CONFIG_MPTCP)
SKB_EXT_MPTCP,
#endif
-#if IS_ENABLED(CONFIG_KCOV)
SKB_EXT_KCOV_HANDLE,
-#endif
SKB_EXT_NUM, /* must be last */
};
+#ifdef CONFIG_SKB_EXTENSIONS
/**
* struct skb_ext - sk_buff extensions
* @refcnt: 1 on allocation, deallocated on 0
@@ -4252,6 +4250,10 @@ static inline void skb_ext_del(struct sk
static inline void __skb_ext_copy(struct sk_buff *d, const struct sk_buff *s) {}
static inline void skb_ext_copy(struct sk_buff *dst, const struct sk_buff *s) {}
static inline bool skb_has_extensions(struct sk_buff *skb) { return false; }
+static inline void *skb_ext_add(struct sk_buff *skb, enum skb_ext_id id)
+{ return NULL; }
+static inline void *skb_ext_find(const struct sk_buff *skb, enum skb_ext_id id)
+{ return NULL; }
#endif /* CONFIG_SKB_EXTENSIONS */
static inline void nf_reset_ct(struct sk_buff *skb)
@@ -4608,7 +4610,6 @@ static inline void skb_reset_redirect(st
#endif
}
-#ifdef CONFIG_KCOV
static inline void skb_set_kcov_handle(struct sk_buff *skb,
const u64 kcov_handle)
{
@@ -4632,11 +4633,6 @@ static inline u64 skb_get_kcov_handle(st
return kcov_handle ? *kcov_handle : 0;
}
-#else
-static inline void skb_set_kcov_handle(struct sk_buff *skb,
- const u64 kcov_handle) { }
-static inline u64 skb_get_kcov_handle(struct sk_buff *skb) { return 0; }
-#endif /* CONFIG_KCOV */
#endif /* __KERNEL__ */
#endif /* _LINUX_SKBUFF_H */
Hi Randy,
On 16/11/2020 04:17, Randy Dunlap wrote:
> The previous Kconfig patch led to some other build errors as
> reported by the 0day bot and my own overnight build testing.
>
> These are all in <linux/skbuff.h> when KCOV is enabled but
> SKB_EXTENSIONS is not enabled, so fix those by combining those conditions
> in the header file.
>
> Also, add stubs for skb_ext_add() and skb_ext_find() to reduce the
> amount of ifdef-ery. (Jakub)
It makes sense, good idea!
Thank you for the new version!
> --- linux-next-20201113.orig/include/linux/skbuff.h
> +++ linux-next-20201113/include/linux/skbuff.h
> @@ -4137,7 +4137,6 @@ static inline void skb_set_nfct(struct s
> #endif
> }
>
> -#ifdef CONFIG_SKB_EXTENSIONS
> enum skb_ext_id {
> #if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
> SKB_EXT_BRIDGE_NF,
> @@ -4151,12 +4150,11 @@ enum skb_ext_id {
> #if IS_ENABLED(CONFIG_MPTCP)
> SKB_EXT_MPTCP,
> #endif
> -#if IS_ENABLED(CONFIG_KCOV)
> SKB_EXT_KCOV_HANDLE,
> -#endif
I don't think we should remove this #ifdef: the number of extensions are
currently limited to 8, we might not want to always have KCOV there even
if we don't want it. I think adding items in this enum only when needed
was the intension of Florian (+cc) when creating these SKB extensions.
Also, this will increase a tiny bit some structures, see "struct skb_ext()".
But apart from that, I think we are fine, even if we add new extensions
in the future after this kcov one.
So if we think it is better to remove these #ifdef here, we should be
OK. But if we prefer not to do that, we should then not add stubs for
skb_ext_{add,find}() and keep the ones for skb_[gs]et_kcov_handle().
Cheers,
Matt
--
Tessares | Belgium | Hybrid Access Solutions
http://www.tessares.net
Matthieu Baerts <[email protected]> wrote:
> > --- linux-next-20201113.orig/include/linux/skbuff.h
> > +++ linux-next-20201113/include/linux/skbuff.h
> > @@ -4137,7 +4137,6 @@ static inline void skb_set_nfct(struct s
> > #endif
> > }
> > -#ifdef CONFIG_SKB_EXTENSIONS
> > enum skb_ext_id {
> > #if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
> > SKB_EXT_BRIDGE_NF,
> > @@ -4151,12 +4150,11 @@ enum skb_ext_id {
> > #if IS_ENABLED(CONFIG_MPTCP)
> > SKB_EXT_MPTCP,
> > #endif
> > -#if IS_ENABLED(CONFIG_KCOV)
> > SKB_EXT_KCOV_HANDLE,
> > -#endif
>
> I don't think we should remove this #ifdef: the number of extensions are
> currently limited to 8, we might not want to always have KCOV there even if
> we don't want it. I think adding items in this enum only when needed was the
> intension of Florian (+cc) when creating these SKB extensions.
> Also, this will increase a tiny bit some structures, see "struct skb_ext()".
Yes, I would also prefer to retrain the ifdef.
Another reason was to make sure that any skb_ext_add(..., MY_EXT) gives
a compile error if the extension is not enabled.
> So if we think it is better to remove these #ifdef here, we should be OK.
> But if we prefer not to do that, we should then not add stubs for
> skb_ext_{add,find}() and keep the ones for skb_[gs]et_kcov_handle().
Yes, exactly, I did not add these stubs because I could not figure out
a case where an empty skb_ext_{add,find} would be wanted.
If your code calls skb_ext_add() but no skb extensions exist you forgot
a SELECT/DEPENDS SKB_EXTENSIONS in Kconfig & compiler error would tell
you that.
On Mon, 16 Nov 2020 15:31:21 +0100 Florian Westphal wrote:
> > > @@ -4151,12 +4150,11 @@ enum skb_ext_id {
> > > #if IS_ENABLED(CONFIG_MPTCP)
> > > SKB_EXT_MPTCP,
> > > #endif
> > > -#if IS_ENABLED(CONFIG_KCOV)
> > > SKB_EXT_KCOV_HANDLE,
> > > -#endif
> >
> > I don't think we should remove this #ifdef: the number of extensions are
> > currently limited to 8, we might not want to always have KCOV there even if
> > we don't want it. I think adding items in this enum only when needed was the
> > intension of Florian (+cc) when creating these SKB extensions.
> > Also, this will increase a tiny bit some structures, see "struct skb_ext()".
>
> Yes, I would also prefer to retrain the ifdef.
>
> Another reason was to make sure that any skb_ext_add(..., MY_EXT) gives
> a compile error if the extension is not enabled.
Oh well, sorry for taking you down the wrong path Randy!
On 11/16/20 7:30 AM, Jakub Kicinski wrote:
> On Mon, 16 Nov 2020 15:31:21 +0100 Florian Westphal wrote:
>>>> @@ -4151,12 +4150,11 @@ enum skb_ext_id {
>>>> #if IS_ENABLED(CONFIG_MPTCP)
>>>> SKB_EXT_MPTCP,
>>>> #endif
>>>> -#if IS_ENABLED(CONFIG_KCOV)
>>>> SKB_EXT_KCOV_HANDLE,
>>>> -#endif
>>>
>>> I don't think we should remove this #ifdef: the number of extensions are
>>> currently limited to 8, we might not want to always have KCOV there even if
>>> we don't want it. I think adding items in this enum only when needed was the
>>> intension of Florian (+cc) when creating these SKB extensions.
>>> Also, this will increase a tiny bit some structures, see "struct skb_ext()".
>>
>> Yes, I would also prefer to retrain the ifdef.
>>
>> Another reason was to make sure that any skb_ext_add(..., MY_EXT) gives
>> a compile error if the extension is not enabled.
>
> Oh well, sorry for taking you down the wrong path Randy!
No problem.
So we are back to v2, right?
Do I need to resend that one?
thanks.
--
~Randy
Randy Dunlap <[email protected]> wrote:
> On 11/16/20 7:30 AM, Jakub Kicinski wrote:
> > On Mon, 16 Nov 2020 15:31:21 +0100 Florian Westphal wrote:
> >>>> @@ -4151,12 +4150,11 @@ enum skb_ext_id {
> >>>> #if IS_ENABLED(CONFIG_MPTCP)
> >>>> SKB_EXT_MPTCP,
> >>>> #endif
> >>>> -#if IS_ENABLED(CONFIG_KCOV)
> >>>> SKB_EXT_KCOV_HANDLE,
> >>>> -#endif
> >>>
> >>> I don't think we should remove this #ifdef: the number of extensions are
> >>> currently limited to 8, we might not want to always have KCOV there even if
> >>> we don't want it. I think adding items in this enum only when needed was the
> >>> intension of Florian (+cc) when creating these SKB extensions.
> >>> Also, this will increase a tiny bit some structures, see "struct skb_ext()".
> >>
> >> Yes, I would also prefer to retrain the ifdef.
> >>
> >> Another reason was to make sure that any skb_ext_add(..., MY_EXT) gives
> >> a compile error if the extension is not enabled.
> >
> > Oh well, sorry for taking you down the wrong path Randy!
>
> No problem.
> So we are back to v2, right?
Yes, you can still drop the line
>> +#if IS_ENABLED(CONFIG_KCOV) && IS_ENABLED(CONFIG_SKB_EXTENSIONS)
for enum skb_ext_id (alreadyt under SKB_EXTENSIONS).
Other than that v2 looks good to me.
Thanks!