Re: [PATCH v4 1/2] netfilter: Introduce new 64-bit helper functions

[Ander Juaristi <a@juaristi.eus>]

[-- Attachment #1: Type: text/plain, Size: 2603 bytes --]



On 13/8/19 20:58, Pablo Neira Ayuso wrote:
> On Tue, Aug 13, 2019 at 08:38:19PM +0200, Ander Juaristi wrote:
> [...]
>> diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h
>> index 9b624566b82d..aa33ada8728a 100644
>> --- a/include/net/netfilter/nf_tables.h
>> +++ b/include/net/netfilter/nf_tables.h
>> @@ -2,6 +2,7 @@
>>  #ifndef _NET_NF_TABLES_H
>>  #define _NET_NF_TABLES_H
>>  
>> +#include <asm/unaligned.h>
>>  #include <linux/list.h>
>>  #include <linux/netfilter.h>
>>  #include <linux/netfilter/nfnetlink.h>
>> @@ -119,6 +120,16 @@ static inline void nft_reg_store8(u32 *dreg, u8 val)
>>  	*(u8 *)dreg = val;
>>  }
>>  
>> +static inline void nft_reg_store64(u32 *dreg, u64 val)
>> +{
>> +	put_unaligned(val, (u64 *)dreg);
>> +}
>> +
>> +static inline u64 nft_reg_load64(u32 *sreg)
>> +{
>> +	return get_unaligned((u64 *)sreg);
>> +}
> 
> Please, add these function definition below _load16() and _store16().

You mean you'd like them ordered from smaller to larger?

nft_reg_store8
nft_reg_load8
nft_reg_store16
nft_reg_load16
nft_reg_store64
nft_reg_load64

>> +
>>  static inline u16 nft_reg_load16(u32 *sreg)
>>  {
>>  	return *(u16 *)sreg;
>> diff --git a/net/netfilter/nft_byteorder.c b/net/netfilter/nft_byteorder.c
>> index e06318428ea0..a25a222d94c8 100644
>> --- a/net/netfilter/nft_byteorder.c
>> +++ b/net/netfilter/nft_byteorder.c
>> @@ -43,14 +43,14 @@ void nft_byteorder_eval(const struct nft_expr *expr,
>>  		switch (priv->op) {
>>  		case NFT_BYTEORDER_NTOH:
> 
> This is network-to-host byteorder.
> 
>>  			for (i = 0; i < priv->len / 8; i++) {
>> -				src64 = get_unaligned((u64 *)&src[i]);
>> -				put_unaligned_be64(src64, &dst[i]);
>> +				src64 = nft_reg_load64(&src[i]);
>> +				nft_reg_store64(&dst[i], cpu_to_be64(src64));
> 
> This looks inverted, this should be:
> 
> 				nft_reg_store64(&dst[i], be64_to_cpu(src64));
> 
> right?
> 
>>  			}
>>  			break;
>>  		case NFT_BYTEORDER_HTON:
> 
> Here, network-to-host byteorder:
> 
>>  			for (i = 0; i < priv->len / 8; i++) {
>> -				src64 = get_unaligned_be64(&src[i]);
>> -				put_unaligned(src64, (u64 *)&dst[i]);
>> +				src64 = be64_to_cpu(nft_reg_load64(&src[i]));
> 
> and this:
> 
>                                 src64 = (__force __u64)
>                                         cpu_to_be64(nft_reg_load64(&src[i]));
> 

My bad. Yes, I've just fixed them.

> The (__force __u64) just makes 'sparse' happy [1].
> 
> [1] https://www.kernel.org/doc/html/v4.12/dev-tools/sparse.html
> 

[-- Attachment #2: pEpkey.asc --]
[-- Type: application/pgp-keys, Size: 6263 bytes --]