[21.02] eal/linux: improve no hugepages logging
diff mbox series

Message ID 11a0d8735a3ee33d02bc57dc702995486051f5d6.1605022892.git.anatoly.burakov@intel.com
State Accepted
Delegated to: David Marchand
Headers show
Series
  • [21.02] eal/linux: improve no hugepages logging
Related show

Checks

Context Check Description
ci/travis-robot success Travis build: passed
ci/checkpatch success coding style OK

Commit Message

Burakov, Anatoly Nov. 10, 2020, 3:41 p.m. UTC
When no hugepages are found, we log a message about it, but we never
specify on which node. We also implicitly declare the page size based
on the directory name, but that's not very user friendly.

Fix both by changing the text of the message to note the NUMA node (if
applicable) and explicitly mention page size in kilobytes.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
 lib/librte_eal/linux/eal_hugepage_info.c | 21 ++++++++++++---------
 1 file changed, 12 insertions(+), 9 deletions(-)

Comments

David Marchand Jan. 12, 2021, 11:17 a.m. UTC | #1
On Tue, Nov 10, 2020 at 4:41 PM Anatoly Burakov
<anatoly.burakov@intel.com> wrote:
>
> When no hugepages are found, we log a message about it, but we never
> specify on which node. We also implicitly declare the page size based
> on the directory name, but that's not very user friendly.
>
> Fix both by changing the text of the message to note the NUMA node (if
> applicable) and explicitly mention page size in kilobytes.

Not sure it is worth aligning but another log uses bytes.
From my tests (after unmounting 2M hugetlbfs mountpoint):
EAL: 512 hugepages of size 2097152 reserved, but no mounted hugetlbfs
found for that size

>
> Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>

The rest lgtm.

Reviewed-by: David Marchand <david.marchand@redhat.com>
David Marchand Jan. 19, 2021, 12:13 p.m. UTC | #2
On Tue, Jan 12, 2021 at 12:17 PM David Marchand
<david.marchand@redhat.com> wrote:
> On Tue, Nov 10, 2020 at 4:41 PM Anatoly Burakov
> <anatoly.burakov@intel.com> wrote:
> >
> > When no hugepages are found, we log a message about it, but we never
> > specify on which node. We also implicitly declare the page size based
> > on the directory name, but that's not very user friendly.
> >
> > Fix both by changing the text of the message to note the NUMA node (if
> > applicable) and explicitly mention page size in kilobytes.
>
> > Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
> Reviewed-by: David Marchand <david.marchand@redhat.com>

Applied.

Patch
diff mbox series

diff --git a/lib/librte_eal/linux/eal_hugepage_info.c b/lib/librte_eal/linux/eal_hugepage_info.c
index 4191af6a21..d97792cade 100644
--- a/lib/librte_eal/linux/eal_hugepage_info.c
+++ b/lib/librte_eal/linux/eal_hugepage_info.c
@@ -84,7 +84,7 @@  static int get_hp_sysfs_value(const char *subdir, const char *file, unsigned lon
 /* this function is only called from eal_hugepage_info_init which itself
  * is only called from a primary process */
 static uint32_t
-get_num_hugepages(const char *subdir)
+get_num_hugepages(const char *subdir, size_t sz)
 {
 	unsigned long resv_pages, num_pages, over_pages, surplus_pages;
 	const char *nr_hp_file = "free_hugepages";
@@ -117,8 +117,8 @@  get_num_hugepages(const char *subdir)
 		over_pages = 0;
 
 	if (num_pages == 0 && over_pages == 0)
-		RTE_LOG(WARNING, EAL, "No available hugepages reported in %s\n",
-				subdir);
+		RTE_LOG(WARNING, EAL, "No available %zu kB hugepages reported\n",
+				sz >> 10);
 
 	num_pages += over_pages;
 	if (num_pages < over_pages) /* overflow */
@@ -133,7 +133,7 @@  get_num_hugepages(const char *subdir)
 }
 
 static uint32_t
-get_num_hugepages_on_node(const char *subdir, unsigned int socket)
+get_num_hugepages_on_node(const char *subdir, unsigned int socket, size_t sz)
 {
 	char path[PATH_MAX], socketpath[PATH_MAX];
 	DIR *socketdir;
@@ -158,8 +158,8 @@  get_num_hugepages_on_node(const char *subdir, unsigned int socket)
 		return 0;
 
 	if (num_pages == 0)
-		RTE_LOG(WARNING, EAL, "No free hugepages reported in %s\n",
-				subdir);
+		RTE_LOG(WARNING, EAL, "No free %zu kB hugepages reported on node %u\n",
+				sz >> 10, socket);
 
 	/*
 	 * we want to return a uint32_t and more than this looks suspicious
@@ -361,7 +361,8 @@  calc_num_pages(struct hugepage_info *hpi, struct dirent *dirent)
 			int socket = rte_socket_id_by_idx(i);
 			unsigned int num_pages =
 					get_num_hugepages_on_node(
-						dirent->d_name, socket);
+						dirent->d_name, socket,
+						hpi->hugepage_sz);
 			hpi->num_pages[socket] = num_pages;
 			total_pages += num_pages;
 		}
@@ -370,7 +371,8 @@  calc_num_pages(struct hugepage_info *hpi, struct dirent *dirent)
 	 * back to old way
 	 */
 	if (total_pages == 0) {
-		hpi->num_pages[0] = get_num_hugepages(dirent->d_name);
+		hpi->num_pages[0] = get_num_hugepages(dirent->d_name,
+				hpi->hugepage_sz);
 
 #ifndef RTE_ARCH_64
 		/* for 32-bit systems, limit number of hugepages to
@@ -418,7 +420,8 @@  hugepage_info_init(void)
 			hpi->hugedir, sizeof(hpi->hugedir)) < 0) {
 			uint32_t num_pages;
 
-			num_pages = get_num_hugepages(dirent->d_name);
+			num_pages = get_num_hugepages(dirent->d_name,
+					hpi->hugepage_sz);
 			if (num_pages > 0)
 				RTE_LOG(NOTICE, EAL,
 					"%" PRIu32 " hugepages of size "