From: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> To: Stephen Hemminger <stephen@networkplumber.org>, dev@dpdk.org Cc: suanmingm@nvidia.com Subject: Re: [dpdk-dev] [PATCH 1/2] ethdev: make flow API primary/secondary process safe Date: Tue, 8 Jun 2021 11:07:13 +0300 Message-ID: <3fd9cfbc-ed10-a11e-f10c-73ea3864acf9@oktetlabs.ru> (raw) In-Reply-To: <20210315192722.35490-2-stephen@networkplumber.org> On 3/15/21 10:27 PM, Stephen Hemminger wrote: > Posix mutex are not by default safe for protecting for usage > from multiple processes. The flow ops mutex could be used by > both primary and secondary processes. > > Bugzilla ID: 662 > Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> > Fixes: 80d1a9aff7f6 ("ethdev: make flow API thread safe") > Cc: suanmingm@nvidia.com > --- > lib/librte_ethdev/rte_ethdev.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/lib/librte_ethdev/rte_ethdev.c b/lib/librte_ethdev/rte_ethdev.c > index 6f514c388b4e..d1024df408a5 100644 > --- a/lib/librte_ethdev/rte_ethdev.c > +++ b/lib/librte_ethdev/rte_ethdev.c > @@ -470,6 +470,7 @@ rte_eth_dev_allocate(const char *name) > { > uint16_t port_id; > struct rte_eth_dev *eth_dev = NULL; > + pthread_mutexattr_t attr; > size_t name_len; > > name_len = strnlen(name, RTE_ETH_NAME_MAX_LEN); > @@ -506,7 +507,10 @@ rte_eth_dev_allocate(const char *name) > strlcpy(eth_dev->data->name, name, sizeof(eth_dev->data->name)); > eth_dev->data->port_id = port_id; > eth_dev->data->mtu = RTE_ETHER_MTU; > - pthread_mutex_init(ð_dev->data->flow_ops_mutex, NULL); > + > + pthread_mutexattr_init(&attr); > + pthread_mutexattr_setpshared(&attr, PTHREAD_PROCESS_SHARED); Return value must be checked here. It may return ENOTSUP and EINVAL. If it fails, IMHO we should do cleanup and return NULL. > + pthread_mutex_init(ð_dev->data->flow_ops_mutex, &attr); > > unlock: > rte_spinlock_unlock(ð_dev_shared_data->ownership_lock); > Please, fix notes from Thomas and above. Overall these patches LGTM. I think we should not introduce any flags that flow ops are multi-process safe. Nobody prevents to call the API in multi-process case and it must behave consistently. The patch makes ethdev API safe and it is responsibility of the driver care about final multi-process safety.
next prev parent reply other threads:[~2021-06-08 8:07 UTC|newest] Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-03-15 19:27 [dpdk-dev] [PATCH 0/2] Mark shared pthread mutex Stephen Hemminger 2021-03-15 19:27 ` [dpdk-dev] [PATCH 1/2] ethdev: make flow API primary/secondary process safe Stephen Hemminger 2021-03-16 23:48 ` Suanming Mou 2021-03-17 0:13 ` Stephen Hemminger 2021-03-17 0:32 ` Suanming Mou 2021-04-14 13:06 ` Ferruh Yigit 2021-04-15 2:55 ` Suanming Mou 2021-04-15 3:17 ` Stephen Hemminger 2021-04-15 7:42 ` Ferruh Yigit 2021-04-15 20:04 ` Stephen Hemminger 2021-04-16 0:57 ` Suanming Mou 2021-04-16 3:19 ` Ajit Khaparde 2021-04-16 1:41 ` fengchengwen 2021-04-16 8:12 ` Ferruh Yigit 2021-04-16 8:18 ` Ferruh Yigit 2021-04-19 17:14 ` Thomas Monjalon 2021-04-19 17:45 ` Stephen Hemminger 2021-04-19 18:09 ` Thomas Monjalon 2021-06-08 8:07 ` Andrew Rybchenko [this message] 2021-03-15 19:27 ` [dpdk-dev] [PATCH 2/2] net/failsafe: fix primary/secondary mutex Stephen Hemminger 2021-04-14 13:10 ` Ferruh Yigit 2021-04-16 8:19 ` Ferruh Yigit 2021-04-19 17:08 ` Thomas Monjalon 2021-06-08 8:00 ` Andrew Rybchenko 2021-06-08 15:42 ` Stephen Hemminger 2021-06-08 15:55 ` Andrew Rybchenko 2021-06-08 20:48 ` Stephen Hemminger 2021-06-09 10:04 ` Andrew Rybchenko 2021-06-14 14:43 ` Gaëtan Rivet 2021-03-15 19:45 ` [dpdk-dev] [PATCH 0/2] Mark shared pthread mutex Stephen Hemminger 2021-03-16 16:28 ` Stephen Hemminger 2021-04-16 8:25 ` 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=3fd9cfbc-ed10-a11e-f10c-73ea3864acf9@oktetlabs.ru \ --to=andrew.rybchenko@oktetlabs.ru \ --cc=dev@dpdk.org \ --cc=stephen@networkplumber.org \ --cc=suanmingm@nvidia.com \ /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
DPDK patches and discussions This inbox may be cloned and mirrored by anyone: git clone --mirror https://inbox.dpdk.org/dev/0 dev/git/0.git # If you have public-inbox 1.1+ installed, you may # initialize and index your mirror using the following commands: public-inbox-init -V2 dev dev/ https://inbox.dpdk.org/dev \ dev@dpdk.org public-inbox-index dev Example config snippet for mirrors. Newsgroup available over NNTP: nntp://inbox.dpdk.org/inbox.dpdk.dev AGPL code for this site: git clone https://public-inbox.org/public-inbox.git