mm/balloon_compaction: Informative allocation warnings
Subsystems (2)
PATCH v2 1/1 mm/balloon_compaction: Informative allocation warnings
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
---
v1->v2:
* Print informative warnings instead suppressing [David]
--- mm/balloon_compaction.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
mm/balloon_compaction.c +6 -1
@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) {struct page *page =alloc_page (balloon_mapping_gfp_mask () | - __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);
Signature · 1 line
2.17.1
Quoted 32 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]--- mm/balloon_compaction.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |
mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);
Not sure if "memory balloon" is the right wording. hmmm.
Signature · 3 lines
Thanks, David / dhildenb
Quoted 32 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]--- mm/balloon_compaction.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |
mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);
Not sure if "memory balloon" is the right wording. hmmm.
Signature · 3 lines
Thanks, David / dhildenb
Quoted 47 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 1 line
Do you have a better suggestion?
Quoted 47 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 1 line
Do you have a better suggestion?
Quoted 47 lines · Nadav Amit
Quoted 47 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 4 lines
Quoted 1 line
Do you have a better suggestion?
Not really - that's why I ack'ed :)
However, thinking about it - what about moving the check + print to the caller and then using dev_warn... or sth. like simple "virtio_balloon:
..." ? You can then drop the warning for vmware balloon if you feel like not needing it.
Signature · 3 lines
Thanks, David / dhildenb
Quoted 47 lines · Nadav Amit
Quoted 47 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 4 lines
Quoted 1 line
Do you have a better suggestion?
Not really - that's why I ack'ed :)
However, thinking about it - what about moving the check + print to the caller and then using dev_warn... or sth. like simple "virtio_balloon:
..." ? You can then drop the warning for vmware balloon if you feel like not needing it.
Signature · 3 lines
Thanks, David / dhildenb
Quoted 52 lines · David Hildenbrand
Quoted 49 lines · Nadav Amit
Quoted 48 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 8 lines
Quoted 1 line
Do you have a better suggestion?
Not really - that's why I ack'ed :)
However, thinking about it - what about moving the check + print to the caller and then using
dev_warn... or sth. like simple "virtio_balloon:
..." ? You can then drop the warning for vmware balloon if you feel like not needing it.
Actually, there is already a warning that is printed by the virtue_balloon in fill_balloon():
struct page *page = balloon_page_alloc();
if (!page) {
dev_info_ratelimited(&vb->vdev->dev,
"Out of puff! Can't get %u pages\n",
VIRTIO_BALLOON_PAGES_PER_PAGE);
/* Sleep for at least 1/5 of a second before retry. */
msleep(200);
break;
}So are you ok with going back to v1?
Quoted 52 lines · David Hildenbrand
Quoted 49 lines · Nadav Amit
Quoted 48 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 8 lines
Quoted 1 line
Do you have a better suggestion?
Not really - that's why I ack'ed :)
However, thinking about it - what about moving the check + print to the caller and then using
dev_warn... or sth. like simple "virtio_balloon:
..." ? You can then drop the warning for vmware balloon if you feel like not needing it.
Actually, there is already a warning that is printed by the virtue_balloon in fill_balloon():
struct page *page = balloon_page_alloc();
if (!page) {
dev_info_ratelimited(&vb->vdev->dev,
"Out of puff! Can't get %u pages\n",
VIRTIO_BALLOON_PAGES_PER_PAGE);
/* Sleep for at least 1/5 of a second before retry. */
msleep(200);
break;
}So are you ok with going back to v1?
Quoted 75 lines · Nadav Amit
Quoted 52 lines · David Hildenbrand
Quoted 49 lines · Nadav Amit
Quoted 48 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 8 lines
Quoted 1 line
Do you have a better suggestion?
Not really - that's why I ack'ed :)
However, thinking about it - what about moving the check + print to the caller and then using
dev_warn... or sth. like simple "virtio_balloon:
..." ? You can then drop the warning for vmware balloon if you feel like not needing it.Actually, there is already a warning that is printed by the
virtue_ballooninfill_balloon():struct page *page = balloon_page_alloc(); if (!page) { dev_info_ratelimited(&vb->vdev->dev, "Out of puff! Can't get %u pages\n", VIRTIO_BALLOON_PAGES_PER_PAGE); /* Sleep for at least 1/5 of a second before retry. */ msleep(200); break; }
Quoted 2 lines
So are you ok with going back to v1?
Whoops, I missed that - sorry - usually the warnings scream louder at me :D
Yes, v1 is fine with me!
Signature · 3 lines
Thanks, David / dhildenb
Quoted 75 lines · Nadav Amit
Quoted 52 lines · David Hildenbrand
Quoted 49 lines · Nadav Amit
Quoted 48 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 8 lines
Quoted 1 line
Do you have a better suggestion?
Not really - that's why I ack'ed :)
However, thinking about it - what about moving the check + print to the caller and then using
dev_warn... or sth. like simple "virtio_balloon:
..." ? You can then drop the warning for vmware balloon if you feel like not needing it.Actually, there is already a warning that is printed by the
virtue_ballooninfill_balloon():struct page *page = balloon_page_alloc(); if (!page) { dev_info_ratelimited(&vb->vdev->dev, "Out of puff! Can't get %u pages\n", VIRTIO_BALLOON_PAGES_PER_PAGE); /* Sleep for at least 1/5 of a second before retry. */ msleep(200); break; }
Quoted 2 lines
So are you ok with going back to v1?
Whoops, I missed that - sorry - usually the warnings scream louder at me :D
Yes, v1 is fine with me!
Signature · 3 lines
Thanks, David / dhildenb
Quoted 78 lines · David Hildenbrand
Quoted 75 lines · Nadav Amit
Quoted 52 lines · David Hildenbrand
Quoted 49 lines · Nadav Amit
Quoted 48 lines · David Hildenbrand
Quoted 31 lines · Nadav Amit
There is no reason to print generic warnings when balloon memory allocation fails, as failures are expected and can be handled gracefully. Since VMware balloon now uses balloon-compaction infrastructure, and suppressed these warnings before, it is also beneficial to suppress these warnings to keep the same behavior that the balloon had before.
Since such warnings can still be useful to indicate that the balloon is over-inflated, print more informative and less frightening warning if allocation fails instead.
Cc David Hildenbrand <[email protected]>Cc Jason Wang <[email protected]>Signed-off-by Nadav Amit <[email protected]>---
v1->v2:
* Print informative warnings instead suppressing [David]
---mm/balloon_compaction.c| 7 ++++++-1 file changed, 6 insertions(+), 1 deletion(-)mm/balloon_compaction.c +0
-0@@ -124,7 +124,12 @@ EXPORT_SYMBOL_GPL(balloon_page_list_dequeue);struct page *balloon_page_alloc (void ) { struct page *page = alloc_page(balloon_mapping_gfp_mask() |mm/balloon_compaction.c EXPORT_SYMBOL_GPL(balloon_page_list_dequeue); · line 135
Quoted 10 lines
mm/balloon_compaction.c +6
-1- __GFP_NOMEMALLOC | __GFP_NORETRY); + __GFP_NOMEMALLOC | __GFP_NORETRY | + __GFP_NOWARN); + +if (!page) +pr_warn_ratelimited ("memory balloon: memory allocation failed" ); +return page; }EXPORT_SYMBOL_GPL (balloon_page_alloc);Not sure if "memory balloon" is the right wording. hmmm.
Quoted 8 lines
Quoted 1 line
Do you have a better suggestion?
Not really - that's why I ack'ed :)
However, thinking about it - what about moving the check + print to the caller and then using
dev_warn... or sth. like simple "virtio_balloon:
..." ? You can then drop the warning for vmware balloon if you feel like not needing it.Actually, there is already a warning that is printed by the
virtue_ballooninfill_balloon():struct page *page = balloon_page_alloc(); if (!page) { dev_info_ratelimited(&vb->vdev->dev, "Out of puff! Can't get %u pages\n", VIRTIO_BALLOON_PAGES_PER_PAGE); /* Sleep for at least 1/5 of a second before retry. */ msleep(200); break; }
Quoted 5 lines
Quoted 1 line
So are you ok with going back to v1?
Whoops, I missed that - sorry - usually the warnings scream louder at me :D
Yes, v1 is fine with me!
Thanks, I missed this one too. This change should prevent making users concerned for no good reason.