From: Anatoly Burakov <anatoly.burakov@intel.com>
To: dev@dpdk.org
Cc: ray.kinsella@intel.com, kuralamudhan.ramakrishnan@intel.com,
louise.m.daly@intel.com, bruce.richardson@intel.com,
ferruh.yigit@intel.com, konstantin.ananyev@intel.com
Subject: [dpdk-dev] [RFC 05/10] mem: add support for no-shared-files mode
Date: Thu, 31 May 2018 15:32:28 +0100 [thread overview]
Message-ID: <968c1d9580b90d4fd7e606cd9e5d838748abdda3.1527776837.git.anatoly.burakov@intel.com> (raw)
In-Reply-To: <cover.1527776837.git.anatoly.burakov@intel.com>
In-Reply-To: <cover.1527776837.git.anatoly.burakov@intel.com>
Unlink hugepages after creating them, to honor the no shared files mode.
We cannot resize non-existing files, so make single file segments
explicitly unsupported.
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
lib/librte_eal/linuxapp/eal/eal.c | 9 +++++++++
lib/librte_eal/linuxapp/eal/eal_memalloc.c | 23 +++++++++++++++++++---
2 files changed, 29 insertions(+), 3 deletions(-)
diff --git a/lib/librte_eal/linuxapp/eal/eal.c b/lib/librte_eal/linuxapp/eal/eal.c
index 32ca25dc2..7904f813e 100644
--- a/lib/librte_eal/linuxapp/eal/eal.c
+++ b/lib/librte_eal/linuxapp/eal/eal.c
@@ -690,6 +690,15 @@ eal_parse_args(int argc, char **argv)
goto out;
}
+ if (internal_config.single_file_segments &&
+ internal_config.no_shared_files) {
+ RTE_LOG(ERR, EAL, "Option --"OPT_SINGLE_FILE_SEGMENTS" is "
+ "incompatible with --"OPT_NO_SHARED_FILES"\n");
+ eal_usage(prgname);
+ ret = -1;
+ goto out;
+ }
+
if (optind >= 0)
argv[optind-1] = prgname;
ret = optind-1;
diff --git a/lib/librte_eal/linuxapp/eal/eal_memalloc.c b/lib/librte_eal/linuxapp/eal/eal_memalloc.c
index 8c11f98c9..f57d307dd 100644
--- a/lib/librte_eal/linuxapp/eal/eal_memalloc.c
+++ b/lib/librte_eal/linuxapp/eal/eal_memalloc.c
@@ -512,6 +512,13 @@ alloc_seg(struct rte_memseg *ms, void *addr, int socket_id,
__func__, strerror(errno));
goto resized;
}
+ if (internal_config.no_shared_files) {
+ if (unlink(path)) {
+ RTE_LOG(DEBUG, EAL, "%s(): unlink() failed: %s\n",
+ __func__, strerror(errno));
+ goto resized;
+ }
+ }
}
/*
@@ -562,8 +569,11 @@ alloc_seg(struct rte_memseg *ms, void *addr, int socket_id,
(unsigned int)(alloc_sz >> 20));
goto mapped;
}
- /* for non-single file segments, we can close fd here */
- if (!internal_config.single_file_segments)
+ /* for non-single file segments or no shared files mode, we can close fd
+ * here
+ */
+ if (!internal_config.single_file_segments ||
+ internal_config.no_shared_files)
close(fd);
/* we need to trigger a write to the page to enforce page fault and
@@ -592,7 +602,8 @@ alloc_seg(struct rte_memseg *ms, void *addr, int socket_id,
/* ignore failure, can't make it any worse */
} else {
/* only remove file if we can take out a write lock */
- if (lock(fd, LOCK_EX) == 1)
+ if (internal_config.no_shared_files == 0 &&
+ lock(fd, LOCK_EX) == 1)
unlink(path);
close(fd);
}
@@ -617,6 +628,12 @@ free_seg(struct rte_memseg *ms, struct hugepage_info *hi,
return -1;
}
+ /* if we're no in shared files mode, nothing needs to be done */
+ if (internal_config.no_shared_files) {
+ memset(ms, 0, sizeof(*ms));
+ return 0;
+ }
+
/* if we are not in single file segments mode, we're going to unmap the
* segment and thus drop the lock on original fd, but hugepage dir is
* now locked so we can take out another one without races.
--
2.17.0
next prev parent reply other threads:[~2018-05-31 14:32 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-05-31 14:32 [dpdk-dev] [RFC 00/10] Support running DPDK without hugetlbfs mountpoint Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 01/10] eal: add --no-shared-files option Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 02/10] eal: make --no-shconf an alias for --no-shared-files Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 03/10] eal: make --huge-unlink " Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 04/10] fbarray: support no-shared-files mode Anatoly Burakov
2018-05-31 14:32 ` Anatoly Burakov [this message]
2018-05-31 14:32 ` [dpdk-dev] [RFC 06/10] ipc: add support for " Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 07/10] eal: add support for no-shared-files for hugepage info Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 08/10] eal: add support for no-shared-files in hugepage data file Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 09/10] eal: do not create runtime dir in no-shared-files mode Anatoly Burakov
2018-05-31 14:32 ` [dpdk-dev] [RFC 10/10] mem: enable memfd-based hugepage allocation Anatoly Burakov
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=968c1d9580b90d4fd7e606cd9e5d838748abdda3.1527776837.git.anatoly.burakov@intel.com \
--to=anatoly.burakov@intel.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=ferruh.yigit@intel.com \
--cc=konstantin.ananyev@intel.com \
--cc=kuralamudhan.ramakrishnan@intel.com \
--cc=louise.m.daly@intel.com \
--cc=ray.kinsella@intel.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
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).