linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Bert Karwatzki <spasswolf@web.de>
To: "Liam R . Howlett" <Liam.Howlett@oracle.com>
Cc: Bert Karwatzki <spasswolf@web.de>,
	Suren Baghdasaryan <surenb@google.com>,
	Vlastimil Babka <vbabka@suse.cz>,
	Lorenzo Stoakes <lstoakes@gmail.com>,
	Matthew Wilcox <willy@infradead.org>,
	sidhartha.kumar@oracle.com,
	"Paul E . McKenney" <paulmck@kernel.org>,
	Jiri Olsa <olsajiri@gmail.com>,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	Andrew Morton <akpm@linux-foundation.org>,
	Kees Cook <kees@kernel.org>, Jeff Xu <jeffxu@chromium.org>,
	"Liam R . Howlett" <Liam.Howlett@Oracle.com>,
	Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Subject: [PATCH v5.1 11/19] mm/mmap: Reposition vma iterator in mmap_region()
Date: Fri, 16 Aug 2024 13:13:52 +0200	[thread overview]
Message-ID: <20240816111405.11793-12-spasswolf@web.de> (raw)
In-Reply-To: <20240816111405.11793-1-spasswolf@web.de>

Instead of moving (or leaving) the vma iterator pointing at the previous
vma, leave it pointing at the insert location.  Pointing the vma
iterator at the insert location allows for a cleaner walk of the vma
tree for MAP_FIXED and the no expansion cases.

The vma_prev() call in the case of merging the previous vma is
equivalent to vma_iter_prev_range(), since the vma iterator will be
pointing to the location just before the previous vma.

Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
---
 mm/mmap.c | 38 ++++++++++++++++++++++----------------
 1 file changed, 22 insertions(+), 16 deletions(-)

diff --git a/mm/mmap.c b/mm/mmap.c
index 16fd14b243f9..6720b55b47ed 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -1404,21 +1404,22 @@ unsigned long mmap_region(struct file *file, unsigned long addr,
 		mas_init(&mas_detach, &mt_detach, /* addr = */ 0);
 		/* Prepare to unmap any existing mapping in the area */
 		if (vms_gather_munmap_vmas(&vms, &mas_detach))
-			return -ENOMEM;
+			goto gather_failed;

 		/* Remove any existing mappings from the vma tree */
 		if (vma_iter_clear_gfp(&vmi, addr, end, GFP_KERNEL))
-			return -ENOMEM;
+			goto clear_tree_failed;

 		/* Unmap any existing mapping in the area */
 		vms_complete_munmap_vmas(&vms, &mas_detach);
 		next = vms.next;
 		prev = vms.prev;
-		vma_prev(&vmi);
 		vma = NULL;
 	} else {
 		next = vma_next(&vmi);
 		prev = vma_prev(&vmi);
+		if (prev)
+			vma_iter_next_range(&vmi);
 	}

 	/*
@@ -1431,11 +1432,8 @@ unsigned long mmap_region(struct file *file, unsigned long addr,
 		vm_flags |= VM_ACCOUNT;
 	}

-	if (vm_flags & VM_SPECIAL) {
-		if (prev)
-			vma_iter_next_range(&vmi);
+	if (vm_flags & VM_SPECIAL)
 		goto cannot_expand;
-	}

 	/* Attempt to expand an old mapping */
 	/* Check next */
@@ -1456,19 +1454,21 @@ unsigned long mmap_region(struct file *file, unsigned long addr,
 		merge_start = prev->vm_start;
 		vma = prev;
 		vm_pgoff = prev->vm_pgoff;
-	} else if (prev) {
-		vma_iter_next_range(&vmi);
+		vma_prev(&vmi); /* Equivalent to going to the previous range */
 	}

-	/* Actually expand, if possible */
-	if (vma &&
-	    !vma_expand(&vmi, vma, merge_start, merge_end, vm_pgoff, next)) {
-		khugepaged_enter_vma(vma, vm_flags);
-		goto expanded;
+	if (vma) {
+		/* Actually expand, if possible */
+		if (!vma_expand(&vmi, vma, merge_start, merge_end, vm_pgoff, next)) {
+			khugepaged_enter_vma(vma, vm_flags);
+			goto expanded;
+		}
+
+		/* If the expand fails, then reposition the vma iterator */
+		if (unlikely(vma == prev))
+			vma_iter_set(&vmi, addr);
 	}

-	if (vma == prev)
-		vma_iter_set(&vmi, addr);
 cannot_expand:

 	/*
@@ -1629,6 +1629,12 @@ unsigned long mmap_region(struct file *file, unsigned long addr,
 		vm_unacct_memory(charged);
 	validate_mm(mm);
 	return error;
+
+clear_tree_failed:
+	abort_munmap_vmas(&mas_detach);
+gather_failed:
+	validate_mm(mm);
+	return -ENOMEM;
 }

 static int __vm_munmap(unsigned long start, size_t len, bool unlock)
--
2.45.2



  parent reply	other threads:[~2024-08-16 11:14 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-16 11:13 [PATCH v5.1 00/19] Rebase v5 patchset to next-20240816 Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 01/19] mm/mmap: Correctly position vma_iterator in __split_vma() Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 02/19] mm/mmap: Introduce abort_munmap_vmas() Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 03/19] mm/mmap: Introduce vmi_complete_munmap_vmas() Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 04/19] mm/mmap: Extract the gathering of vmas from do_vmi_align_munmap() Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 05/19] mm/mmap: Introduce vma_munmap_struct for use in munmap operations Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 06/19] mm/mmap: Change munmap to use vma_munmap_struct() for accounting and surrounding vmas Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 07/19] mm/mmap: Extract validate_mm() from vma_complete() Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 08/19] mm/mmap: Inline munmap operation in mmap_region() Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 09/19] mm/mmap: Expand mmap_region() munmap call Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 10/19] mm/mmap: Support vma == NULL in init_vma_munmap() Bert Karwatzki
2024-08-16 11:13 ` Bert Karwatzki [this message]
2024-08-16 11:13 ` [PATCH v5.1 12/19] mm/mmap: Track start and end of munmap in vma_munmap_struct Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 13/19] mm/mmap: Clean up unmap_region() argument list Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 14/19] mm/mmap: Avoid zeroing vma tree in mmap_region() Bert Karwatzki
2024-08-16 11:13 ` [PATCH v5.1 15/19] mm/mmap: Use PHYS_PFN " Bert Karwatzki
2024-08-16 11:33 ` [PATCH v5.1 00/19] Rebase v5 patchset to next-20240816 Lorenzo Stoakes
2024-08-16 11:55   ` Bert Karwatzki

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=20240816111405.11793-12-spasswolf@web.de \
    --to=spasswolf@web.de \
    --cc=Liam.Howlett@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=jeffxu@chromium.org \
    --cc=kees@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lorenzo.stoakes@oracle.com \
    --cc=lstoakes@gmail.com \
    --cc=olsajiri@gmail.com \
    --cc=paulmck@kernel.org \
    --cc=sidhartha.kumar@oracle.com \
    --cc=surenb@google.com \
    --cc=vbabka@suse.cz \
    --cc=willy@infradead.org \
    /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