RDMA/nldev: Add copy-on-fork attribute to get sys command
PATCH v4 1/1 RDMA/nldev: Add copy-on-fork attribute to get sys command
The new attribute indicates that the kernel copies DMA pages on fork, hence libibverbs' fork support through madvise and MADV_DONTFORK is not needed.
The introduced attribute is always reported as supported since the kernel has the patch that added the copy-on-fork behavior. This allows the userspace library to identify older vs newer kernel versions. Extra care should be taken when backporting this patch as it relies on the fact that the copy-on-fork patch is merged, hence no check for support is added.
Don't backport this patch unless you also have the following series:70e806e4e645 ("mm: Do early cow for pinned pages during fork() for ptes") and 4eae4efa2c29 ("hugetlb: do early cow when page pinned on src mm").
---
PR was sent:
https://github.com/linux-rdma/rdma-core/pull/975
Changelog -v3->v4: https://lore.kernel.org/linux-rdma/[email protected]/
* Mention that nla_put_u8() return value is ignored on purpose
v2->v3: https://lore.kernel.org/linux-rdma/[email protected]/
* Remove check if copy-on-fork attribute was provided from nldev_set_sys_set_doit()
v1->v2: https://lore.kernel.org/linux-rdma/[email protected]/
* Remove nla_put_u8() return value check
* Add commit hashes to commit message and code comment
--- drivers/infiniband/core/nldev.c | 14 ++++++++++++++ include/uapi/rdma/rdma_netlink.h | 2 ++ 2 files changed, 16 insertions(+)
drivers/infiniband/core/nldev.c +14 -0
@@ -146,6 +146,7 @@ static const struct nla_policy nldev_policy[RDMA_NLDEV_ATTR_MAX] = { [RDMA_NLDEV_ATTR_UVERBS_DRIVER_ID] = { .type = NLA_U32 }, [RDMA_NLDEV_NET_NS_FD] = { .type = NLA_U32 }, [RDMA_NLDEV_SYS_ATTR_NETNS_MODE] = { .type = NLA_U8 }, + [RDMA_NLDEV_SYS_ATTR_COPY_ON_FORK] = { .type = NLA_U8 }, }; static int put_driver_name_print_type(struct sk_buff *msg, const char *name,
@@ -1697,6 +1698,19 @@ static int nldev_sys_get_doit(struct sk_buff *skb, struct nlmsghdr *nlh,nlmsg_free (msg);return err; } + +/* +* Copy-on-fork is supported. +* See commits: +* 70e806e4e645 ("mm: Do early cow for pinned pages during fork() for ptes") +* 4eae4efa2c29 ("hugetlb: do early cow when page pinned on src mm") +* for more details. Don't backport this without them. +* +* Return value ignored on purpose, assume copy-on-fork is not +* supported in case of failure. +*/ +nla_put_u8 (msg,RDMA_NLDEV_SYS_ATTR_COPY_ON_FORK ,1 ); +nlmsg_end (msg, nlh);return rdma_nl_unicast (sock_net (skb->sk), msg,NETLINK_CB (skb).portid); }
include/uapi/rdma/rdma_netlink.h +2 -0
@@ -533,6 +533,8 @@ enum rdma_nldev_attr {RDMA_NLDEV_ATTR_RES_RAW ,/* binary */ + RDMA_NLDEV_SYS_ATTR_COPY_ON_FORK, /* u8 */ +/* * Always the end */
base-commit: 7d8f346504ebde71d92905e3055d40ea8f34416e
Signature · 1 line
2.31.1
Quoted 30 lines · Gal Pressman
The new attribute indicates that the kernel copies DMA pages on fork, hence libibverbs' fork support through madvise and
MADV_DONTFORKis not needed.The introduced attribute is always reported as supported since the kernel has the patch that added the copy-on-fork behavior. This allows the userspace library to identify older vs newer kernel versions. Extra care should be taken when backporting this patch as it relies on the fact that the copy-on-fork patch is merged, hence no check for support is added.
Don't backport this patch unless you also have the following series:
70e806e4e645("mm: Do early cow for pinned pages duringfork()for ptes") and4eae4efa2c29("hugetlb: do early cow when page pinned on src mm").Fixes 70e806e4e645 ("mm: Do early cow for pinned pages during fork() for ptes")Fixes 4eae4efa2c29 ("hugetlb: do early cow when page pinned on src mm")Signed-off-by Gal Pressman <[email protected]>Reviewed-by Leon Romanovsky <[email protected]>---
PR was sent:
https://github.com/linux-rdma/rdma-core/pull/975Changelog -
v3->v4: https://lore.kernel.org/linux-rdma/[email protected]/
* Mention thatnla_put_u8()return value is ignored on purpose
v2->v3: https://lore.kernel.org/linux-rdma/[email protected]/
* Remove check if copy-on-fork attribute was provided fromnldev_set_sys_set_doit()
Quoted 7 lines
v1->v2: https://lore.kernel.org/linux-rdma/[email protected]/
* Removenla_put_u8()return value check
* Add commit hashes to commit message and code comment--- drivers/infiniband/core/nldev.c | 14 ++++++++++++++ include/uapi/rdma/rdma_netlink.h | 2 ++ 2 files changed, 16 insertions(+)
Applied to for-next, not the enum number will have changed compared to what you sent
Thanks,
Jason