acl: Fix RTE_ACL_RULE_SZ macro definition
diff mbox series

Message ID 20200707122643.29222-1-levendsayar@gmail.com
State Rejected
Delegated to: Thomas Monjalon
Headers show
Series
  • acl: Fix RTE_ACL_RULE_SZ macro definition
Related show

Checks

Context Check Description
ci/iol-intel-Performance success Performance Testing PASS
ci/iol-mellanox-Performance success Performance Testing PASS
ci/Intel-compilation success Compilation OK
ci/iol-testing success Testing PASS
ci/travis-robot success Travis build: passed
ci/iol-broadcom-Performance success Performance Testing PASS
ci/checkpatch success coding style OK

Commit Message

Levend Sayar July 7, 2020, 12:26 p.m. UTC
From: Levend Sayar <levendsayar@gmail.com>

Signed-off-by: Levend Sayar <levendsayar@gmail.com>
---
 lib/librte_acl/rte_acl.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Levend Sayar July 7, 2020, 3:25 p.m. UTC | #1
Sure.
I am really sorry for not being verbose enough.

From lib/librte_acl/rte_acl.h

#define RTE_ACL_RULE_DEF(name, fld_num) struct name {\
    struct rte_acl_rule_data data;               \
    struct rte_acl_field field[fld_num];         \
}

RTE_ACL_RULE_DEF(rte_acl_rule,);

When you put the definition in-place, above line means :

struct rte_acl_rule {
   struct rte_acl_rule_data data;
   struct rte_acl_field field[];
}

There is another define to get the size of an acl rule such as

#define RTE_ACL_RULE_SZ(fld_num)    \
    (sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) * (fld_num))

So the above definition gets the size of a "struct rte_acl_rule" which has
fld_num fields.
which must be
*sizeof (struct rte_acl_rule_data) *+ (sizeof(struct rte_acl_field) *
fld_num)

Because it adds up the sizes of struct components;

But according to the current RTE_ACL_RULE_SZ, it is

*sizeof (struct rte_acl_rule)* + (sizeof(struct rte_acl_field) * fld_num)

So my patch only changes the part that I underlined.

sizeof (struct rte_acl_rule) = 16;
sizeof (struct rte_acl_rule_data) = 12;

Best,
Levend


On Tue, Jul 7, 2020 at 3:42 PM Ananyev, Konstantin <
konstantin.ananyev@intel.com> wrote:

>
> > From: Levend Sayar <levendsayar@gmail.com>
>
> Could you provide some explanation:
> What do you think is wrong with current version and why,
> and what your fix does.
>
> >
> > Signed-off-by: Levend Sayar <levendsayar@gmail.com>
> > ---
> >  lib/librte_acl/rte_acl.h | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/lib/librte_acl/rte_acl.h b/lib/librte_acl/rte_acl.h
> > index aa22e70c6..d34fdbc0e 100644
> > --- a/lib/librte_acl/rte_acl.h
> > +++ b/lib/librte_acl/rte_acl.h
> > @@ -116,7 +116,7 @@ struct rte_acl_rule_data {
> >  RTE_ACL_RULE_DEF(rte_acl_rule,);
> >
> >  #define      RTE_ACL_RULE_SZ(fld_num)        \
> > -     (sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) *
> (fld_num))
> > +     (sizeof(struct rte_acl_rule_data) + sizeof(struct rte_acl_field) *
> (fld_num))
> >
> >
> >  /** Max number of characters in name.*/
> > --
> > 2.27.0
>
>
Ananyev, Konstantin July 7, 2020, 4:44 p.m. UTC | #2
From: Levend Sayar <levendsayar@gmail.com>
Sent: Tuesday, July 7, 2020 4:25 PM
To: Ananyev, Konstantin <konstantin.ananyev@intel.com>
Cc: dev@dpdk.org
Subject: Re: [PATCH] acl: Fix RTE_ACL_RULE_SZ macro definition

Sure.
I am really sorry for not being verbose enough.

From lib/librte_acl/rte_acl.h

#define RTE_ACL_RULE_DEF(name, fld_num) struct name {\
    struct rte_acl_rule_data data;               \
    struct rte_acl_field field[fld_num];         \
}

RTE_ACL_RULE_DEF(rte_acl_rule,);

When you put the definition in-place, above line means :

struct rte_acl_rule {
   struct rte_acl_rule_data data;
   struct rte_acl_field field[];
}

[KA] Yes.

There is another define to get the size of an acl rule such as

#define RTE_ACL_RULE_SZ(fld_num)    \
    (sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) * (fld_num))

So the above definition gets the size of a "struct rte_acl_rule" which has fld_num fields.
which must be
sizeof (struct rte_acl_rule_data) + (sizeof(struct rte_acl_field) * fld_num)

Because it adds up the sizes of struct components;

[KA] I don’t think so.
You forgot about possible gaps between members of rte_acl_rule.
Let say for 64 bit target it would be a 4B gap between ‘data’ and ‘field’.
So, for:
RTE_ACL_RULE_DEF(xz, 1);

sizeof(struct xz) == RTE_ACL_RULE_SZ(1) == 32

After changes you suggest
RTE_ACL_RULE_SZ(1) == 28 != sizeof(struct xz)
Which is wrong.

But according to the current RTE_ACL_RULE_SZ, it is

sizeof (struct rte_acl_rule) + (sizeof(struct rte_acl_field) * fld_num)

So my patch only changes the part that I underlined.

sizeof (struct rte_acl_rule) = 16;
sizeof (struct rte_acl_rule_data) = 12;

Best,
Levend


On Tue, Jul 7, 2020 at 3:42 PM Ananyev, Konstantin <konstantin.ananyev@intel.com<mailto:konstantin.ananyev@intel.com>> wrote:

> From: Levend Sayar <levendsayar@gmail.com<mailto:levendsayar@gmail.com>>

Could you provide some explanation:
What do you think is wrong with current version and why,
and what your fix does.

>
> Signed-off-by: Levend Sayar <levendsayar@gmail.com<mailto:levendsayar@gmail.com>>
> ---
>  lib/librte_acl/rte_acl.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/librte_acl/rte_acl.h b/lib/librte_acl/rte_acl.h
> index aa22e70c6..d34fdbc0e 100644
> --- a/lib/librte_acl/rte_acl.h
> +++ b/lib/librte_acl/rte_acl.h
> @@ -116,7 +116,7 @@ struct rte_acl_rule_data {
>  RTE_ACL_RULE_DEF(rte_acl_rule,);
>
>  #define      RTE_ACL_RULE_SZ(fld_num)        \
> -     (sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) * (fld_num))
> +     (sizeof(struct rte_acl_rule_data) + sizeof(struct rte_acl_field) * (fld_num))
>
>
>  /** Max number of characters in name.*/
> --
> 2.27.0
Levend Sayar July 7, 2020, 4:59 p.m. UTC | #3
Yes you are totally right.
I overlooked the alignment of struct components.
Sorry for that.

But imho, not to have such confusions,

Another define can be added such as

#define RTE_ACL_RULE_SIZE(xz) sizeof(struct xz)

Thanks for your time.

Best,
Levend




On Tue, Jul 7, 2020 at 7:45 PM Ananyev, Konstantin <
konstantin.ananyev@intel.com> wrote:

>
>
>
>
> *From:* Levend Sayar <levendsayar@gmail.com>
> *Sent:* Tuesday, July 7, 2020 4:25 PM
> *To:* Ananyev, Konstantin <konstantin.ananyev@intel.com>
> *Cc:* dev@dpdk.org
> *Subject:* Re: [PATCH] acl: Fix RTE_ACL_RULE_SZ macro definition
>
>
>
> Sure.
>
> I am really sorry for not being verbose enough.
>
>
>
> From lib/librte_acl/rte_acl.h
>
>
>
> #define RTE_ACL_RULE_DEF(*name*, *fld_num*) *struct* name {\
>
>     *struct* rte_acl_rule_data data;               \
>
>     *struct* rte_acl_field field[fld_num];         \
>
> }
>
>
>
> RTE_ACL_RULE_DEF(rte_acl_rule,);
>
>
>
> When you put the definition in-place, above line means :
>
>
>
> struct rte_acl_rule {
>
>    struct rte_acl_rule_data data;
>
>    struct rte_acl_field field[];
>
> }
>
>
>
> [KA] Yes.
>
>
>
> There is another define to get the size of an acl rule such as
>
>
>
> #define RTE_ACL_RULE_SZ(*fld_num*)    \
>
>     (sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) * (fld_num
> ))
>
>
>
> So the above definition gets the size of a "struct rte_acl_rule" which has
> fld_num fields.
>
> which must be
>
> *sizeof (struct rte_acl_rule_data) *+ (sizeof(struct rte_acl_field) *
> fld_num)
>
>
>
> Because it adds up the sizes of struct components;
>
>
>
> [KA] I don’t think so.
>
> You forgot about possible gaps between members of rte_acl_rule.
>
> Let say for 64 bit target it would be a 4B gap between ‘data’ and ‘field’.
>
> So, for:
>
> RTE_ACL_RULE_DEF(xz, 1);
>
>
>
> sizeof(struct xz) == RTE_ACL_RULE_SZ(1) == 32
>
>
>
> After changes you suggest
>
> RTE_ACL_RULE_SZ(1) == 28 != sizeof(struct xz)
>
> Which is wrong.
>
>
>
> But according to the current RTE_ACL_RULE_SZ, it is
>
>
>
> *sizeof (struct rte_acl_rule)* + (sizeof(struct rte_acl_field) * fld_num)
>
>
>
> So my patch only changes the part that I underlined.
>
>
>
> sizeof (struct rte_acl_rule) = 16;
>
> sizeof (struct rte_acl_rule_data) = 12;
>
>
>
> Best,
>
> Levend
>
>
>
>
>
> On Tue, Jul 7, 2020 at 3:42 PM Ananyev, Konstantin <
> konstantin.ananyev@intel.com> wrote:
>
>
> > From: Levend Sayar <levendsayar@gmail.com>
>
> Could you provide some explanation:
> What do you think is wrong with current version and why,
> and what your fix does.
>
> >
> > Signed-off-by: Levend Sayar <levendsayar@gmail.com>
> > ---
> >  lib/librte_acl/rte_acl.h | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/lib/librte_acl/rte_acl.h b/lib/librte_acl/rte_acl.h
> > index aa22e70c6..d34fdbc0e 100644
> > --- a/lib/librte_acl/rte_acl.h
> > +++ b/lib/librte_acl/rte_acl.h
> > @@ -116,7 +116,7 @@ struct rte_acl_rule_data {
> >  RTE_ACL_RULE_DEF(rte_acl_rule,);
> >
> >  #define      RTE_ACL_RULE_SZ(fld_num)        \
> > -     (sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) *
> (fld_num))
> > +     (sizeof(struct rte_acl_rule_data) + sizeof(struct rte_acl_field) *
> (fld_num))
> >
> >
> >  /** Max number of characters in name.*/
> > --
> > 2.27.0
>
>

Patch
diff mbox series

diff --git a/lib/librte_acl/rte_acl.h b/lib/librte_acl/rte_acl.h
index aa22e70c6..d34fdbc0e 100644
--- a/lib/librte_acl/rte_acl.h
+++ b/lib/librte_acl/rte_acl.h
@@ -116,7 +116,7 @@  struct rte_acl_rule_data {
 RTE_ACL_RULE_DEF(rte_acl_rule,);
 
 #define	RTE_ACL_RULE_SZ(fld_num)	\
-	(sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) * (fld_num))
+	(sizeof(struct rte_acl_rule_data) + sizeof(struct rte_acl_field) * (fld_num))
 
 
 /** Max number of characters in name.*/