From: fengchengwen <fengchengwen@huawei.com>
To: "Morten Brørup" <mb@smartsharesystems.com>,
thomas@monjalon.net, ferruh.yigit@amd.com
Cc: <dev@dpdk.org>, <huangdengdui@huawei.com>,
<stephen@networkplumber.org>, <roretzla@linux.microsoft.com>
Subject: Re: [PATCH] ethdev: fix strict aliasing lead to link cannot be up
Date: Thu, 11 Apr 2024 19:57:37 +0800 [thread overview]
Message-ID: <2987eb7f-5439-9ecc-494c-b968f059da50@huawei.com> (raw)
In-Reply-To: <98CBD80474FA8B44BF855DF32C47DC35E9F394@smartserver.smartshare.dk>
Hi Morten,
On 2024/4/11 14:58, Morten Brørup wrote:
>> From: Chengwen Feng [mailto:fengchengwen@huawei.com]
>> Sent: Thursday, 11 April 2024 05.08
>>
>> Fix a problem introduced by a compiler upgrade (from gcc10 to gcc12.3),
>> which will lead the hns3 NIC can't link up. The root cause is strict
>> aliasing violation in rte_eth_linkstatus_set() with hns3 driver, see
>> [1] for more details.
>>
>> This commit use union to avoid such aliasing violation.
>>
>> [1] Strict aliasing problem with rte_eth_linkstatus_set()
>> https://marc.info/?l=dpdk-dev&m=171274148514777&w=3
>>
>> Cc: stable@dpdk.org
>>
>> Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
>> Signed-off-by: Dengdui Huang <huangdengdui@huawei.com>
>> ---
>
> The patch mixes atomic and non-atomic access.
> This is not new for DPDK, which used to rely on compiler built-in atomics.
>
> I'm not sure it needs to be changed, but my suggestion is inline below.
> I don't think it makes any practical different for 64 bit arch, but it might for 32 bit arch.
>
>> lib/ethdev/ethdev_driver.h | 23 +++++++----------------
>> lib/ethdev/rte_ethdev.h | 16 ++++++++++------
>> 2 files changed, 17 insertions(+), 22 deletions(-)
>>
>> diff --git a/lib/ethdev/ethdev_driver.h b/lib/ethdev/ethdev_driver.h
>> index 0dbf2dd6a2..9d831d5c84 100644
>> --- a/lib/ethdev/ethdev_driver.h
>> +++ b/lib/ethdev/ethdev_driver.h
>> @@ -1674,18 +1674,13 @@ static inline int
>> rte_eth_linkstatus_set(struct rte_eth_dev *dev,
>> const struct rte_eth_link *new_link)
>> {
>> - RTE_ATOMIC(uint64_t) *dev_link = (uint64_t __rte_atomic *)&(dev-
>>> data->dev_link);
>> - union {
>> - uint64_t val64;
>> - struct rte_eth_link link;
>> - } orig;
>> -
>> - RTE_BUILD_BUG_ON(sizeof(*new_link) != sizeof(uint64_t));
>> + struct rte_eth_link old_link;
>>
>> - orig.val64 = rte_atomic_exchange_explicit(dev_link, *(const
>> uint64_t *)new_link,
>> - rte_memory_order_seq_cst);
>> + old_link.val64 = rte_atomic_exchange_explicit(&dev->data-
>>> dev_link.val64,
>
> old_link.val64 should be written using:
> rte_atomic_store_explicit(&old_link.val64, ..., rte_memory_order_seq_cst)
I'm afraid I don't agree this, the &dev->data->dev_link.val64 should use atomic not the stack variable old_link.
>
>> + new_link->val64,
>
> new_link->val64 should be read using:
> rte_atomic_load_explicit(&new_link->val64, rte_memory_order_seq_cst)
The same reason with above.
>
>> + rte_memory_order_seq_cst);
>
>>
>> - return (orig.link.link_status == new_link->link_status) ? -1 : 0;
>> + return (old_link.link_status == new_link->link_status) ? -1 : 0;
>> }
>>
>> /**
>> @@ -1701,12 +1696,8 @@ static inline void
>> rte_eth_linkstatus_get(const struct rte_eth_dev *dev,
>> struct rte_eth_link *link)
>> {
>> - RTE_ATOMIC(uint64_t) *src = (uint64_t __rte_atomic *)&(dev->data-
>>> dev_link);
>> - uint64_t *dst = (uint64_t *)link;
>> -
>> - RTE_BUILD_BUG_ON(sizeof(*link) != sizeof(uint64_t));
>> -
>> - *dst = rte_atomic_load_explicit(src, rte_memory_order_seq_cst);
>> + link->val64 = rte_atomic_load_explicit(&dev->data->dev_link.val64,
>
> link->val64 should be written using:
> rte_atomic_store_explicit(&link->val64, ..., rte_memory_order_seq_cst)
The same reason with above, the &dev->data->dev_link.val64 should use atomic not the stack variable link.
>
>> + rte_memory_order_seq_cst);
>> }
>>
>> /**
>> diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
>> index 147257d6a2..0b5d3d2318 100644
>> --- a/lib/ethdev/rte_ethdev.h
>> +++ b/lib/ethdev/rte_ethdev.h
>> @@ -332,12 +332,16 @@ struct rte_eth_stats {
>> /**
>> * A structure used to retrieve link-level information of an Ethernet
>> port.
>> */
>> -__extension__
>> -struct __rte_aligned(8) rte_eth_link { /**< aligned for atomic64
>> read/write */
>> - uint32_t link_speed; /**< RTE_ETH_SPEED_NUM_ */
>> - uint16_t link_duplex : 1; /**< RTE_ETH_LINK_[HALF/FULL]_DUPLEX
>> */
>> - uint16_t link_autoneg : 1; /**< RTE_ETH_LINK_[AUTONEG/FIXED] */
>> - uint16_t link_status : 1; /**< RTE_ETH_LINK_[DOWN/UP] */
>> +struct rte_eth_link {
>> + union {
>> + uint64_t val64; /**< used for atomic64 read/write */
>
> The type of val64 should be:
> RTE_ATOMIC(uint64_t)
ack
Plus: yes, this patch mixes atomic and non-atomic access, but the main reason
is that we want to simplify the implementation. If we want to separate it clearly,
maybe we should defined as this:
struct rte_eth_link {
union {
RTE_ATOMIC(uint64_t) atomic64; /**< used for atomic64 read/write */
struct {
uint64_t val64;
};
struct {
uint32_t link_speed; /**< RTE_ETH_SPEED_NUM_ */
uint16_t link_duplex : 1; /**< RTE_ETH_LINK_[HALF/FULL]_DUPLEX */
uint16_t link_autoneg : 1; /**< RTE_ETH_LINK_[AUTONEG/FIXED] */
uint16_t link_status : 1; /**< RTE_ETH_LINK_[DOWN/UP] */
};
};
};
Thanks
>
>> + struct {
>> + uint32_t link_speed; /**< RTE_ETH_SPEED_NUM_
>> */
>> + uint16_t link_duplex : 1; /**<
>> RTE_ETH_LINK_[HALF/FULL]_DUPLEX */
>> + uint16_t link_autoneg : 1; /**<
>> RTE_ETH_LINK_[AUTONEG/FIXED] */
>> + uint16_t link_status : 1; /**<
>> RTE_ETH_LINK_[DOWN/UP] */
>> + };
>> + };
>> };
>>
>> /**@{@name Link negotiation
>> --
>> 2.17.1
>
> .
>
next prev parent reply other threads:[~2024-04-11 11:57 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-11 3:07 Chengwen Feng
2024-04-11 6:53 ` Morten Brørup
2024-04-11 6:58 ` Morten Brørup
2024-04-11 11:57 ` fengchengwen [this message]
2024-04-11 12:44 ` Morten Brørup
2024-04-11 12:04 ` [PATCH v3] " Chengwen Feng
2024-04-11 12:44 ` Morten Brørup
2024-04-12 3:27 ` fengchengwen
2024-04-12 7:24 ` Morten Brørup
2024-04-11 15:05 ` [PATCH] " Stephen Hemminger
2024-04-12 8:16 ` fengchengwen
2024-04-12 8:49 ` [PATCH v4] " Chengwen Feng
2024-04-13 8:04 ` [PATCH v5] " Chengwen Feng
2024-04-13 8:48 ` [PATCH v6] " Chengwen Feng
2024-04-15 13:15 ` Morten Brørup
2024-04-18 7:28 ` [PATCH v7] " Chengwen Feng
2024-04-19 15:15 ` Ferruh Yigit
2024-04-19 15:25 ` Ferruh Yigit
2024-04-22 6:42 ` fengchengwen
2024-04-22 6:38 ` [PATCH v8] " Chengwen Feng
2024-04-22 10:54 ` Ferruh Yigit
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2987eb7f-5439-9ecc-494c-b968f059da50@huawei.com \
--to=fengchengwen@huawei.com \
--cc=dev@dpdk.org \
--cc=ferruh.yigit@amd.com \
--cc=huangdengdui@huawei.com \
--cc=mb@smartsharesystems.com \
--cc=roretzla@linux.microsoft.com \
--cc=stephen@networkplumber.org \
--cc=thomas@monjalon.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).