All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jan Beulich <jbeulich@suse.com>
To: "xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>
Cc: Paul Durrant <paul@xen.org>, Andrew Cooper <andrew.cooper3@citrix.com>
Subject: [PATCH 1/9] AMD/IOMMU: redo awaiting of command completion
Date: Wed, 9 Jun 2021 11:26:48 +0200	[thread overview]
Message-ID: <1bb49cce-537e-9fba-80bb-82bf502728d0@suse.com> (raw)
In-Reply-To: <03285055-47ff-ab10-ae76-0553f28f136d@suse.com>

The present abuse of the completion interrupt does not only stand in the
way of, down the road, using it for its actual purpose, but also
requires holding the IOMMU lock while waiting for command completion,
limiting parallelism and keeping interrupts off for non-negligible
periods of time. Have the IOMMU do an ordinary memory write instead of
signaling an otherwise disabled interrupt (by just updating a status
register bit).

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>

--- a/xen/drivers/passthrough/amd/iommu_cmd.c
+++ b/xen/drivers/passthrough/amd/iommu_cmd.c
@@ -20,6 +20,9 @@
 #include "iommu.h"
 #include "../ats.h"
 
+#define CMD_COMPLETION_INIT 0
+#define CMD_COMPLETION_DONE 1
+
 static void send_iommu_command(struct amd_iommu *iommu,
                                const uint32_t cmd[4])
 {
@@ -49,28 +52,31 @@ static void send_iommu_command(struct am
 static void flush_command_buffer(struct amd_iommu *iommu,
                                  unsigned int timeout_base)
 {
+    static DEFINE_PER_CPU(uint64_t, poll_slot);
+    uint64_t *this_poll_slot = &this_cpu(poll_slot);
+    paddr_t addr = virt_to_maddr(this_poll_slot);
     uint32_t cmd[4];
     s_time_t start, timeout;
     static unsigned int __read_mostly threshold = 1;
 
-    /* RW1C 'ComWaitInt' in status register */
-    writel(IOMMU_STATUS_COMP_WAIT_INT,
-           iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET);
-
-    /* send an empty COMPLETION_WAIT command to flush command buffer */
-    cmd[3] = cmd[2] = 0;
-    set_field_in_reg_u32(IOMMU_CMD_COMPLETION_WAIT, 0,
+    ACCESS_ONCE(*this_poll_slot) = CMD_COMPLETION_INIT;
+
+    /* send a COMPLETION_WAIT command to flush command buffer */
+    cmd[0] = addr;
+    set_field_in_reg_u32(IOMMU_CONTROL_ENABLED, cmd[0],
+                         IOMMU_COMP_WAIT_S_FLAG_MASK,
+                         IOMMU_COMP_WAIT_S_FLAG_SHIFT, &cmd[0]);
+    cmd[1] = addr >> 32;
+    set_field_in_reg_u32(IOMMU_CMD_COMPLETION_WAIT, cmd[1],
                          IOMMU_CMD_OPCODE_MASK,
                          IOMMU_CMD_OPCODE_SHIFT, &cmd[1]);
-    set_field_in_reg_u32(IOMMU_CONTROL_ENABLED, 0,
-                         IOMMU_COMP_WAIT_I_FLAG_MASK,
-                         IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]);
+    cmd[2] = CMD_COMPLETION_DONE;
+    cmd[3] = 0;
     send_iommu_command(iommu, cmd);
 
     start = NOW();
     timeout = start + (timeout_base ?: 100) * MILLISECS(threshold);
-    while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) &
-              IOMMU_STATUS_COMP_WAIT_INT) )
+    while ( ACCESS_ONCE(*this_poll_slot) != CMD_COMPLETION_DONE )
     {
         if ( timeout && NOW() > timeout )
         {



  reply	other threads:[~2021-06-09  9:27 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-09  9:25 [PATCH 0/9] IOMMU: XSA-373 follow-on Jan Beulich
2021-06-09  9:26 ` Jan Beulich [this message]
2021-06-09 10:36   ` [PATCH 1/9] AMD/IOMMU: redo awaiting of command completion Andrew Cooper
2021-06-09 12:08     ` Jan Beulich
2021-06-09 12:33       ` Andrew Cooper
2021-06-09 12:51         ` Jan Beulich
2021-06-10 12:24     ` Jan Beulich
2021-06-09  9:27 ` [PATCH 2/9] AMD/IOMMU: re-work locking around sending of commands Jan Beulich
2021-06-09 10:53   ` Andrew Cooper
2021-06-09 12:22     ` Jan Beulich
2021-06-10 11:58     ` Jan Beulich
2021-06-10 12:53       ` Jan Beulich
2021-06-09  9:27 ` [PATCH 3/9] VT-d: undo device mappings upon error Jan Beulich
2021-06-24  5:13   ` Tian, Kevin
2021-06-09  9:28 ` [PATCH 4/9] VT-d: adjust domid map updating when unmapping context Jan Beulich
2021-06-24  5:21   ` Tian, Kevin
2021-06-09  9:28 ` [PATCH 5/9] VT-d: clear_fault_bits() should clear all fault bits Jan Beulich
2021-06-24  5:26   ` Tian, Kevin
2021-06-09  9:29 ` [PATCH 6/9] VT-d: don't lose errors when flushing TLBs on multiple IOMMUs Jan Beulich
2021-06-24  5:28   ` Tian, Kevin
2021-06-09  9:29 ` [PATCH 7/9] VT-d: centralize mapping of QI entries Jan Beulich
2021-06-24  5:31   ` Tian, Kevin
2021-06-09  9:29 ` [PATCH 8/9] VT-d: drop/move a few QI related constants Jan Beulich
2021-06-24  5:32   ` Tian, Kevin
2021-06-09  9:30 ` [PATCH 9/9] IOMMU/PCI: don't let domain cleanup continue when device de-assignment failed Jan Beulich
2021-06-24 15:34   ` Paul Durrant
2021-06-23  6:51 ` Ping: [PATCH 0/9] IOMMU: XSA-373 follow-on Jan Beulich
2021-06-23  6:58   ` Tian, Kevin

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=1bb49cce-537e-9fba-80bb-82bf502728d0@suse.com \
    --to=jbeulich@suse.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=paul@xen.org \
    --cc=xen-devel@lists.xenproject.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.