[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <6cf8217d-1336-1778-6a2c-61ffdf0ae00a@novek.ru>
Date: Wed, 3 Feb 2021 01:09:28 +0000
From: Vadim Fedorenko <vfedorenko@...ek.ru>
To: Rong Chen <rong.a.chen@...el.com>,
Jakub Kicinski <kuba@...nel.org>,
Boris Pismenny <borisp@...dia.com>,
Aviad Yehezkel <aviadye@...dia.com>
Cc: netdev@...r.kernel.org, kernel test robot <oliver.sang@...el.com>
Subject: Re: [PATCH] selftests/tls: fix compile errors after adding
CHACHA20-POLY1305
On 03.02.2021 00:58, Rong Chen wrote:
>
>
> On 2/2/21 6:11 PM, Vadim Fedorenko wrote:
>> On 02.02.2021 09:45, Rong Chen wrote:
>>> The kernel test robot reported the following errors:
>>>
>>> tls.c: In function ‘tls_setup’:
>>> tls.c:136:27: error: storage size of ‘tls12’ isn’t known
>>> union tls_crypto_context tls12;
>>> ^~~~~
>>> tls.c:150:21: error: ‘tls12_crypto_info_chacha20_poly1305’ undeclared (first
>>> use in this function)
>>> tls12_sz = sizeof(tls12_crypto_info_chacha20_poly1305);
>>> ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>> tls.c:150:21: note: each undeclared identifier is reported only once for each
>>> function it appears in
>>> tls.c:153:21: error: ‘tls12_crypto_info_aes_gcm_128’ undeclared (first use in
>>> this function)
>>> tls12_sz = sizeof(tls12_crypto_info_aes_gcm_128);
>>>
>>> Fixes: 4f336e88a870 ("selftests/tls: add CHACHA20-POLY1305 to tls selftests")
>>> Reported-by: kernel test robot <oliver.sang@...el.com>
>>> Link: https://lore.kernel.org/lkml/20210108064141.GB3437@xsang-OptiPlex-9020/
>>> Signed-off-by: Rong Chen <rong.a.chen@...el.com>
>>> ---
>>> include/net/tls.h | 9 ---------
>>> include/uapi/linux/tls.h | 9 +++++++++
>>> tools/testing/selftests/net/tls.c | 4 ++--
>>> 3 files changed, 11 insertions(+), 11 deletions(-)
>>>
>>> diff --git a/include/net/tls.h b/include/net/tls.h
>>> index 3eccb525e8f7..54f7863ad915 100644
>>> --- a/include/net/tls.h
>>> +++ b/include/net/tls.h
>>> @@ -212,15 +212,6 @@ struct cipher_context {
>>> char *rec_seq;
>>> };
>>> -union tls_crypto_context {
>>> - struct tls_crypto_info info;
>>> - union {
>>> - struct tls12_crypto_info_aes_gcm_128 aes_gcm_128;
>>> - struct tls12_crypto_info_aes_gcm_256 aes_gcm_256;
>>> - struct tls12_crypto_info_chacha20_poly1305 chacha20_poly1305;
>>> - };
>>> -};
>>> -
>>> struct tls_prot_info {
>>> u16 version;
>>> u16 cipher_type;
>>> diff --git a/include/uapi/linux/tls.h b/include/uapi/linux/tls.h
>>> index 0d54baea1d8d..9933dd425571 100644
>>> --- a/include/uapi/linux/tls.h
>>> +++ b/include/uapi/linux/tls.h
>>> @@ -124,6 +124,15 @@ struct tls12_crypto_info_chacha20_poly1305 {
>>> unsigned char rec_seq[TLS_CIPHER_CHACHA20_POLY1305_REC_SEQ_SIZE];
>>> };
>>> +union tls_crypto_context {
>>> + struct tls_crypto_info info;
>>> + union {
>>> + struct tls12_crypto_info_aes_gcm_128 aes_gcm_128;
>>> + struct tls12_crypto_info_aes_gcm_256 aes_gcm_256;
>>> + struct tls12_crypto_info_chacha20_poly1305 chacha20_poly1305;
>>> + };
>>> +};
>>> +
>>> enum {
>>> TLS_INFO_UNSPEC,
>>> TLS_INFO_VERSION,
>>> diff --git a/tools/testing/selftests/net/tls.c
>>> b/tools/testing/selftests/net/tls.c
>>> index e0088c2d38a5..6951c8524a27 100644
>>> --- a/tools/testing/selftests/net/tls.c
>>> +++ b/tools/testing/selftests/net/tls.c
>>> @@ -147,10 +147,10 @@ FIXTURE_SETUP(tls)
>>> tls12.info.cipher_type = variant->cipher_type;
>>> switch (variant->cipher_type) {
>>> case TLS_CIPHER_CHACHA20_POLY1305:
>>> - tls12_sz = sizeof(tls12_crypto_info_chacha20_poly1305);
>>> + tls12_sz = sizeof(struct tls12_crypto_info_chacha20_poly1305);
>>> break;
>>> case TLS_CIPHER_AES_GCM_128:
>>> - tls12_sz = sizeof(tls12_crypto_info_aes_gcm_128);
>>> + tls12_sz = sizeof(struct tls12_crypto_info_aes_gcm_128);
>>> break;
>>> default:
>>> tls12_sz = 0;
>>>
>>
>> I'm not sure that it's a good idea to move tls_crypto_context to uapi as it's
>> an internal union. Previous patches assumes that user-space uses different
>> structures for different cipher configurations.
>
> Hi Vadim,
>
> Sorry i don't have the background, could you help to fix these build errors?
>
> Best Regards,
> Rong Chen
>
Yes, sure, no problem. I will prepare new patch tomorrow.
Powered by blists - more mailing lists