qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Eric Auger <eric.auger@redhat.com>
To: eric.auger.pro@gmail.com, eric.auger@redhat.com,
	qemu-devel@nongnu.org, qemu-arm@nongnu.org, stefanha@redhat.com,
	fam@euphon.net, philmd@redhat.com, alex.williamson@redhat.com
Cc: lvivier@redhat.com, kwolf@redhat.com, cohuck@redhat.com,
	mreitz@redhat.com
Subject: [PATCH v2 2/2] util/vfio-helpers: Rework the IOVA allocator to avoid IOVA reserved regions
Date: Tue, 29 Sep 2020 10:55:50 +0200	[thread overview]
Message-ID: <20200929085550.30926-3-eric.auger@redhat.com> (raw)
In-Reply-To: <20200929085550.30926-1-eric.auger@redhat.com>

Introduce the qemu_vfio_find_fixed/temp_iova helpers which
respectively allocate IOVAs from the bottom/top parts of the
usable IOVA range, without picking within host IOVA reserved
windows. The allocation remains basic: if the size is too big
for the remaining of the current usable IOVA range, we jump
to the next one, leaving a hole in the address map.

Signed-off-by: Eric Auger <eric.auger@redhat.com>

---

v1 -> v2:
- handle possible wrap
---
 util/vfio-helpers.c | 57 +++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 53 insertions(+), 4 deletions(-)

diff --git a/util/vfio-helpers.c b/util/vfio-helpers.c
index ba0ee6e21c..71145970f3 100644
--- a/util/vfio-helpers.c
+++ b/util/vfio-helpers.c
@@ -667,6 +667,50 @@ static bool qemu_vfio_verify_mappings(QEMUVFIOState *s)
     return true;
 }
 
+static int
+qemu_vfio_find_fixed_iova(QEMUVFIOState *s, size_t size, uint64_t *iova)
+{
+    int i;
+
+    for (i = 0; i < s->nb_iova_ranges; i++) {
+        if (s->usable_iova_ranges[i].end < s->low_water_mark) {
+            continue;
+        }
+        s->low_water_mark =
+            MAX(s->low_water_mark, s->usable_iova_ranges[i].start);
+
+        if (s->usable_iova_ranges[i].end - s->low_water_mark + 1 >= size ||
+            s->usable_iova_ranges[i].end - s->low_water_mark + 1 == 0) {
+            *iova = s->low_water_mark;
+            s->low_water_mark += size;
+            return 0;
+        }
+    }
+    return -ENOMEM;
+}
+
+static int
+qemu_vfio_find_temp_iova(QEMUVFIOState *s, size_t size, uint64_t *iova)
+{
+    int i;
+
+    for (i = s->nb_iova_ranges - 1; i >= 0; i--) {
+        if (s->usable_iova_ranges[i].start > s->high_water_mark) {
+            continue;
+        }
+        s->high_water_mark =
+            MIN(s->high_water_mark, s->usable_iova_ranges[i].end + 1);
+
+        if (s->high_water_mark - s->usable_iova_ranges[i].start + 1 >= size ||
+            s->high_water_mark - s->usable_iova_ranges[i].start + 1 == 0) {
+            *iova = s->high_water_mark - size;
+            s->high_water_mark = *iova;
+            return 0;
+        }
+    }
+    return -ENOMEM;
+}
+
 /* Map [host, host + size) area into a contiguous IOVA address space, and store
  * the result in @iova if not NULL. The caller need to make sure the area is
  * aligned to page size, and mustn't overlap with existing mapping areas (split
@@ -693,7 +737,11 @@ int qemu_vfio_dma_map(QEMUVFIOState *s, void *host, size_t size,
             goto out;
         }
         if (!temporary) {
-            iova0 = s->low_water_mark;
+            if (qemu_vfio_find_fixed_iova(s, size, &iova0)) {
+                ret = -ENOMEM;
+                goto out;
+            }
+
             mapping = qemu_vfio_add_mapping(s, host, size, index + 1, iova0);
             if (!mapping) {
                 ret = -ENOMEM;
@@ -705,15 +753,16 @@ int qemu_vfio_dma_map(QEMUVFIOState *s, void *host, size_t size,
                 qemu_vfio_undo_mapping(s, mapping, NULL);
                 goto out;
             }
-            s->low_water_mark += size;
             qemu_vfio_dump_mappings(s);
         } else {
-            iova0 = s->high_water_mark - size;
+            if (qemu_vfio_find_temp_iova(s, size, &iova0)) {
+                ret = -ENOMEM;
+                goto out;
+            }
             ret = qemu_vfio_do_mapping(s, host, size, iova0);
             if (ret) {
                 goto out;
             }
-            s->high_water_mark -= size;
         }
     }
     if (iova) {
-- 
2.21.3



  parent reply	other threads:[~2020-09-29  8:59 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-29  8:55 [PATCH v2 0/2] NVMe passthrough: Take into account host IOVA reserved regions Eric Auger
2020-09-29  8:55 ` [PATCH v2 1/2] util/vfio-helpers: Collect " Eric Auger
2020-09-29  8:55 ` Eric Auger [this message]
2020-09-29 15:59   ` [PATCH v2 2/2] util/vfio-helpers: Rework the IOVA allocator to avoid " Stefan Hajnoczi
2020-09-29 19:44     ` Auger Eric
2020-09-30  9:18       ` Stefan Hajnoczi
2020-09-30  9:23 ` [PATCH v2 0/2] NVMe passthrough: Take into account host " Stefan Hajnoczi

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=20200929085550.30926-3-eric.auger@redhat.com \
    --to=eric.auger@redhat.com \
    --cc=alex.williamson@redhat.com \
    --cc=cohuck@redhat.com \
    --cc=eric.auger.pro@gmail.com \
    --cc=fam@euphon.net \
    --cc=kwolf@redhat.com \
    --cc=lvivier@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=philmd@redhat.com \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@redhat.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).