Skip to contentSkip to conversation

man2: udpate mm/userfaultfd manpages to latest

PATCH Superseded Peter Xu 3 participants 13 messages Mar 4, 2021

Subsystems (2)

v1 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Peter Xu

author
#0 lore raw

cover man2: udpate mm/userfaultfd manpages to latest

There're two features missing in current manpage, namely:

(1) Userfaultfd Thread-ID feature
(2) Userfaultfd write protect mode

There's also a 3rd one which was just contributed from Axel - Axel, I think it would be great if you can add that part too, probably after the whole hugetlbfs/shmem minor mode reaches the linux master branch.

Please review, thanks.

Peter Xu (4):
userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs userfaultfd.2: Add write-protect mode ioctl_userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs ioctl_userfaultfd.2: Add write-protect mode docs

 man2/ioctl_userfaultfd.2 |  76 +++++++++++++++++++++++++++--
 man2/userfaultfd.2       | 100 ++++++++++++++++++++++++++++++++++++++-
 2 files changed, 171 insertions(+), 5 deletions(-)
Signature · 1 line
2.26.2
author
#1 lore raw

PATCH 1/4 userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs

UFFD_FEATURE_THREAD_ID is supported since Linux 4.14.

---
 man2/userfaultfd.2 | 12 ++++++++++++
 1 file changed, 12 insertions(+)
man2/userfaultfd.2 +12 -0
index e7dc9f813..2d14effc6 100644
@@ -77,6 +77,12 @@ When the last file descriptor referring to a userfaultfd object is closed,
 all memory ranges that were registered with the object are unregistered
 and unread events are flushed.
 .\"
+.PP
+Since Linux 4.14, userfaultfd page fault message can selectively embed fault
+thread ID information into the fault message.  One needs to enable this feature
+explicitly using the
+.BR UFFD_FEATURE_THREAD_ID
+feature bit when initializing the userfaultfd context, otherwise disabled.
 .SS Usage
 The userfaultfd mechanism is designed to allow a thread in a multithreaded
 program to perform user-space paging for the other threads in the process.
@@ -229,6 +235,9 @@ struct uffd_msg {
         struct {
             __u64 flags;    /* Flags describing fault */
             __u64 address;  /* Faulting address */
+            union {
+                __u32 ptid; /* Thread ID of the fault */
+            } feat;
         } pagefault;
 
         struct {            /* Since Linux 4.11 */
@@ -358,6 +367,9 @@ otherwise it is a read fault.
 .\" UFFD_PAGEFAULT_FLAG_WP is not yet supported.
 .RE
 .TP
+.I pagefault.feat.pid
+The thread ID that triggered the page fault.
+.TP
 .I fork.ufd
 The file descriptor associated with the userfault object
 created for the child created by
Signature · 1 line
2.26.2
#2 lore raw

Re: [PATCH 1/4] userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs

Quoted 12 lines · Peter Xu

UFFD_FEATURE_THREAD_ID is supported since Linux 4.14.

---
 man2/userfaultfd.2 | 12 ++++++++++++
 1 file changed, 12 insertions(+)
man2/userfaultfd.2 +0 -0
index e7dc9f813..2d14effc6 100644

man2/userfaultfd.2 When the last file descriptor referring to a userfaultfd object is closed, · line 81

Quoted 5 lines
man2/userfaultfd.2 +2 -0
@@ -77,6 +77,12 @@ When the last file descriptor referring to a userfaultfd object is closed,
 all memory ranges that were registered with the object are unregistered
 and unread events are flushed.
 .\"
+.PP
+Since Linux 4.14, userfaultfd page fault message can selectively embed fault

Maybe faulting? ^

man2/userfaultfd.2 When the last file descriptor referring to a userfaultfd object is closed, · line 85

Quoted 4 lines
man2/userfaultfd.2 +4 -0
+thread ID information into the fault message.  One needs to enable this feature
+explicitly using the
+.BR UFFD_FEATURE_THREAD_ID
+feature bit when initializing the userfaultfd context, otherwise disabled.

"... otherwise thread ID reporting is disabled" ^

Quoted 26 lines
man2/userfaultfd.2 +6 -0
 .SS Usage
 The userfaultfd mechanism is designed to allow a thread in a multithreaded
 program to perform user-space paging for the other threads in the process.
@@ -229,6 +235,9 @@ struct uffd_msg {
         struct {
             __u64 flags;    /* Flags describing fault */
             __u64 address;  /* Faulting address */
+            union {
+                __u32 ptid; /* Thread ID of the fault */
+            } feat;
         } pagefault;
 
         struct {            /* Since Linux 4.11 */
@@ -358,6 +367,9 @@ otherwise it is a read fault.
 .\" UFFD_PAGEFAULT_FLAG_WP is not yet supported.
 .RE
 .TP
+.I pagefault.feat.pid
+The thread ID that triggered the page fault.
+.TP
 .I fork.ufd
 The file descriptor associated with the userfault object
 created for the child created by
Signature · 1 line
2.26.2
Signature · 2 lines
Sincerely yours,
Mike.
author
#3 lore raw

Re: [PATCH 1/4] userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs

Quoted 8 lines · Mike Rapoport
Quoted 7 lines · Peter Xu

UFFD_FEATURE_THREAD_ID is supported since Linux 4.14.

---
 man2/userfaultfd.2 | 12 ++++++++++++
 1 file changed, 12 insertions(+)
Quoted 12 lines

man2/userfaultfd.2 When the last file descriptor referring to a userfaultfd object is closed, · line 81

Quoted 10 lines
man2/userfaultfd.2 +2 -0
index e7dc9f813..2d14effc6 100644
@@ -77,6 +77,12 @@ When the last file descriptor referring to a userfaultfd object is closed,
 all memory ranges that were registered with the object are unregistered
 and unread events are flushed.
 .\"
+.PP
+Since Linux 4.14, userfaultfd page fault message can selectively embed fault

Maybe faulting? ^

Ok.

Quoted 7 lines

man2/userfaultfd.2 When the last file descriptor referring to a userfaultfd object is closed, · line 85

Quoted 5 lines
man2/userfaultfd.2 +4 -0
+thread ID information into the fault message.  One needs to enable this feature
+explicitly using the
+.BR UFFD_FEATURE_THREAD_ID
+feature bit when initializing the userfaultfd context, otherwise disabled.

"... otherwise thread ID reporting is disabled" ^

I changed to "... By default, thread ID reporting is diabled."

Thanks!

Signature · 1 line
Peter Xu
#4 lore raw

Re: [PATCH 1/4] userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs

In reply to Peter Xu

Hello Peter,

Quoted 12 lines · Peter Xu

UFFD_FEATURE_THREAD_ID is supported since Linux 4.14.

---
  man2/userfaultfd.2 | 12 ++++++++++++
  1 file changed, 12 insertions(+)
man2/userfaultfd.2 +0 -0
index e7dc9f813..2d14effc6 100644

man2/userfaultfd.2 When the last file descriptor referring to a userfaultfd object is closed, · line 84

Quoted 8 lines
man2/userfaultfd.2 +5 -0
@@ -77,6 +77,12 @@ When the last file descriptor referring to a userfaultfd object is closed,
  all memory ranges that were registered with the object are unregistered
  and unread events are flushed.
  .\"
+.PP
+Since Linux 4.14, userfaultfd page fault message can selectively embed fault
+thread ID information into the fault message.  One needs to enable this feature
+explicitly using the
+.BR UFFD_FEATURE_THREAD_ID

This should use [.B] and not [.BR].
.BR is for alternate Bold and Roman.
.B is for bold.

(There are more appearances of this in the other patches.)

Thanks,

Alex

Signature · 3 lines
Alejandro Colomar
Linux man-pages comaintainer; https://www.kernel.org/doc/man-pages/
http://www.alejandro-colomar.es/
author
#5 lore raw

Re: [PATCH 1/4] userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs

Quoted 1 line · Alejandro Colomar (man-pages)

Hello Peter,

Hi, Alex,

Quoted 27 lines · Peter Xu
Quoted 12 lines

UFFD_FEATURE_THREAD_ID is supported since Linux 4.14.

---
  man2/userfaultfd.2 | 12 ++++++++++++
  1 file changed, 12 insertions(+)
man2/userfaultfd.2 +0 -0
index e7dc9f813..2d14effc6 100644

man2/userfaultfd.2 When the last file descriptor referring to a userfaultfd object is closed, · line 84

Quoted 8 lines
man2/userfaultfd.2 +5 -0
@@ -77,6 +77,12 @@ When the last file descriptor referring to a userfaultfd object is closed,
  all memory ranges that were registered with the object are unregistered
  and unread events are flushed.
  .\"
+.PP
+Since Linux 4.14, userfaultfd page fault message can selectively embed fault
+thread ID information into the fault message.  One needs to enable this feature
+explicitly using the
+.BR UFFD_FEATURE_THREAD_ID

This should use [.B] and not [.BR].
.BR is for alternate Bold and Roman.
.B is for bold.

Quoted 1 line

(There are more appearances of this in the other patches.)

Yeah I got a bit confused when differenciating those two, since I also see similar usage, e.g.:

.BR O_CLOEXEC

I'll fix all of them appeared in current patchset. Let me know if you also want me to fix all the existing uses of ".BR" too where ".B" would suffice. Otherwise I won't touch them since I can't say they're wrong either (I think most of them should generate the same output with either ".BR" or ".B" if there's only one word?).

Thanks,

Signature · 1 line
Peter Xu
#6 lore raw

Re: [PATCH 1/4] userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs

Hi Peter,

Quoted 7 lines · Peter Xu
Quoted 6 lines · Alejandro Colomar (man-pages)
Quoted 1 line

+.BR UFFD_FEATURE_THREAD_ID

This should use [.B] and not [.BR].
.BR is for alternate Bold and Roman.
.B is for bold.

Quoted 6 lines
Quoted 1 line

(There are more appearances of this in the other patches.)

Yeah I got a bit confused when differenciating those two, since I also see similar usage, e.g.:

.BR O_CLOEXEC

Yes, these are minor imperfections that got into the manual pages, and we don't remove them due to the churn that it would create (and possibility of introducing other bugs while doing such a big scripted change that couldn't be easily reviewed (thousands of lines)). So as we still have those lines, they tend to confuse.

Quoted 6 lines

I'll fix all of them appeared in current patchset. Let me know if you also want me to fix all the existing uses of ".BR" too where ".B" would suffice. Otherwise I won't touch them since I can't say they're wrong either (I think most of them should generate the same output with either ".BR" or ".B" if there's only one word?).

Our current non-written guidelines are:
We are fixing the existing ones as we modify code near them,
but leave untouched code that is far from what we are changing, even on the same page.

Thanks,

Alex

Signature · 3 lines
Alejandro Colomar
Linux man-pages comaintainer; https://www.kernel.org/doc/man-pages/
http://www.alejandro-colomar.es/
author
#7 lore raw

PATCH 2/4 userfaultfd.2: Add write-protect mode

In reply to Peter Xu

Write-protect mode is supported starting from Linux 5.7.

---
 man2/userfaultfd.2 | 88 ++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 86 insertions(+), 2 deletions(-)
man2/userfaultfd.2 +86 -2
index 2d14effc6..8e1602d62 100644
@@ -78,6 +78,28 @@ all memory ranges that were registered with the object are unregistered
 and unread events are flushed.
 .\"
 .PP
+Currently, userfaultfd supports two modes of registration:
+.TP
+.BR UFFDIO_REGISTER_MODE_MISSING
+When registered with
+.BR UFFDIO_REGISTER_MODE_MISSING
+mode, the userspace will receive a page fault message when a missing page is
+accessed.  The faulted thread will be stopped from execution until the page
+fault is resolved from the userspace by either an
+.BR UFFDIO_COPY
+or an
+.BR UFFDIO_ZEROPAGE
+ioctl.
+.TP
+.BR UFFDIO_REGISTER_MODE_WP
+When registered with
+.BR UFFDIO_REGISTER_MODE_WP
+mode, the userspace will receive a page fault message when a write-protected
+page is written.  The faulted thread will be stopped from execution until the
+userspace un-write-protect the page using an
+.BR UFFDIO_WRITEPROTECT
+ioctl.
+.PP
 Since Linux 4.14, userfaultfd page fault message can selectively embed fault
 thread ID information into the fault message.  One needs to enable this feature
 explicitly using the
@@ -143,6 +165,16 @@ single threaded non-cooperative userfaultfd manager implementations.
 .\" and limitations remaining in 4.11
 .\" Maybe it's worth adding a dedicated sub-section...
 .\"
+.PP
+Starting from Linux 5.7, userfaultfd is able to do synchronous page dirty
+tracking using the new write-protection register mode.  One should check
+against the feature bit
+.B UFFD_FEATURE_PAGEFAULT_FLAG_WP
+before using this feature.  Similar to the original userfaultfd missing mode,
+the write-protect mode will generate an userfaultfd message when the protected
+page is written.  The user needs to resolve the page fault by unprotecting the
+faulted page and kick the faulted thread to continue.  For more information,
+please read the "Userfaultfd write-protect mode" section below.
 .SS Userfaultfd operation
 After the userfaultfd object is created with
 .BR userfaultfd (),
@@ -218,6 +250,54 @@ userfaultfd can be used only with anonymous private memory mappings.
 Since Linux 4.11,
 userfaultfd can be also used with hugetlbfs and shared memory mappings.
 .\"
+.SS Userfaultfd write-protect mode
+Since Linux 5.7, userfaultfd started to support write-protect mode.  The user
+needs to first check availability of this feature using
+.BR UFFDIO_API
+ioctl against the feature bit
+.BR UFFD_FEATURE_PAGEFAULT_FLAG_WP .
+.PP
+To register with userfaultfd write-protect mode, the user needs to send the
+.BR UFFDIO_REGISTER
+ioctl with mode
+.BR UFFDIO_REGISTER_MODE_WP
+set.  Note that it's legal to monitor the same memory range with multiple
+modes.  For example, the user can do
+.BR UFFDIO_REGISTER
+with the mode set to
+.BR UFFDIO_REGISTER_MODE_MISSING\ |\ UFFDIO_REGISTER_MODE_WP.
+When there is only
+.BR UFFDIO_REGISTER_MODE_WP
+registered, the userspace will
+.I not
+receive any message when a missing page is written.  Instead, the userspace
+will only receive a write-protect page fault message when an existing but
+write-protected page got written.
+.PP
+After the
+.BR UFFDIO_REGISTER
+ioctl completed with
+.BR UFFDIO_REGISTER_MODE_WP
+mode set, one can write-protect any existing memory within the range using the
+ioctl
+.BR UFFDIO_WRITEPROTECT
+where
+.I uffdio_writeprotect.mode
+should be set to
+.BR UFFDIO_WRITEPROTECT_MODE_WP .
+.PP
+When a write-protect event happens, the userspace will receive a page fault
+message whose
+.I uffd_msg.pagefault.flags
+will be with
+.BR UFFD_PAGEFAULT_FLAG_WP
+flag set.  Note: since only writes can trigger such kind of fault,
+write-protect messages will always be with
+.BR UFFD_PAGEFAULT_FLAG_WRITE
+bit set too along with
+.BR UFFD_PAGEFAULT_FLAG_WP .
+.PP
+Currently, write-protect mode only supports private anonymous memory.
 .SS Reading from the userfaultfd structure
 Each
 .BR read (2)
@@ -363,8 +443,12 @@ flag (see
 .BR ioctl_userfaultfd (2))
 and this flag is set, this a write fault;
 otherwise it is a read fault.
-.\"
-.\" UFFD_PAGEFAULT_FLAG_WP is not yet supported.
+.TP
+.B UFFD_PAGEFAULT_FLAG_WP
+If the address is in a range that was registered with the
+.B UFFDIO_REGISTER_MODE_WP
+flag, when this bit is set it means it's a write-protect fault.  Otherwise it's
+a page missing fault.
 .RE
 .TP
 .I pagefault.feat.pid
Signature · 1 line
2.26.2
#8 lore raw

Re: [PATCH 2/4] userfaultfd.2: Add write-protect mode

Quoted 27 lines · Peter Xu

Write-protect mode is supported starting from Linux 5.7.

---
 man2/userfaultfd.2 | 88 ++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 86 insertions(+), 2 deletions(-)
man2/userfaultfd.2 +12 -0
index 2d14effc6..8e1602d62 100644
@@ -78,6 +78,28 @@ all memory ranges that were registered with the object are unregistered
 and unread events are flushed.
 .\"
 .PP
+Currently, userfaultfd supports two modes of registration:
+.TP
+.BR UFFDIO_REGISTER_MODE_MISSING
+When registered with
+.BR UFFDIO_REGISTER_MODE_MISSING
+mode, the userspace will receive a page fault message when a missing page is
+accessed.  The faulted thread will be stopped from execution until the page
+fault is resolved from the userspace by either an
+.BR UFFDIO_COPY
+or an
+.BR UFFDIO_ZEROPAGE
+ioctl.

man2/userfaultfd.2 all memory ranges that were registered with the object are unregistered · line 102

Quoted 10 lines
man2/userfaultfd.2 +10 -0
+.TP
+.BR UFFDIO_REGISTER_MODE_WP
+When registered with
+.BR UFFDIO_REGISTER_MODE_WP
+mode, the userspace will receive a page fault message when a write-protected
+page is written.  The faulted thread will be stopped from execution until the
+userspace un-write-protect the page using an
+.BR UFFDIO_WRITEPROTECT
+ioctl.
+.PP

I'd add a sentence about combining the modes together. Something like

"Both modes can be enabled together for the same memory range"

Quoted 21 lines
man2/userfaultfd.2 +10 -0
 Since Linux 4.14, userfaultfd page fault message can selectively embed fault
 thread ID information into the fault message.  One needs to enable this feature
 explicitly using the
@@ -143,6 +165,16 @@ single threaded non-cooperative userfaultfd manager implementations.
 .\" and limitations remaining in 4.11
 .\" Maybe it's worth adding a dedicated sub-section...
 .\"
+.PP
+Starting from Linux 5.7, userfaultfd is able to do synchronous page dirty
+tracking using the new write-protection register mode.  One should check
+against the feature bit
+.B UFFD_FEATURE_PAGEFAULT_FLAG_WP
+before using this feature.  Similar to the original userfaultfd missing mode,
+the write-protect mode will generate an userfaultfd message when the protected
+page is written.  The user needs to resolve the page fault by unprotecting the
+faulted page and kick the faulted thread to continue.  For more information,
+please read the "Userfaultfd write-protect mode" section below.
 .SS Userfaultfd operation
 After the userfaultfd object is created with
 .BR userfaultfd (),

man2/userfaultfd.2 userfaultfd can be used only with anonymous private memory mappings. · line 254

Quoted 5 lines
man2/userfaultfd.2 +2 -0
@@ -218,6 +250,54 @@ userfaultfd can be used only with anonymous private memory mappings.
 Since Linux 4.11,
 userfaultfd can be also used with hugetlbfs and shared memory mappings.
 .\"
+.SS Userfaultfd write-protect mode
+Since Linux 5.7, userfaultfd started to support write-protect mode.  The user

Maybe s/started to support/supports/

Quoted 67 lines
man2/userfaultfd.2 +52 -2
+needs to first check availability of this feature using
+.BR UFFDIO_API
+ioctl against the feature bit
+.BR UFFD_FEATURE_PAGEFAULT_FLAG_WP .
+.PP
+To register with userfaultfd write-protect mode, the user needs to send the
+.BR UFFDIO_REGISTER
+ioctl with mode
+.BR UFFDIO_REGISTER_MODE_WP
+set.  Note that it's legal to monitor the same memory range with multiple
+modes.  For example, the user can do
+.BR UFFDIO_REGISTER
+with the mode set to
+.BR UFFDIO_REGISTER_MODE_MISSING\ |\ UFFDIO_REGISTER_MODE_WP.
+When there is only
+.BR UFFDIO_REGISTER_MODE_WP
+registered, the userspace will
+.I not
+receive any message when a missing page is written.  Instead, the userspace
+will only receive a write-protect page fault message when an existing but
+write-protected page got written.
+.PP
+After the
+.BR UFFDIO_REGISTER
+ioctl completed with
+.BR UFFDIO_REGISTER_MODE_WP
+mode set, one can write-protect any existing memory within the range using the
+ioctl
+.BR UFFDIO_WRITEPROTECT
+where
+.I uffdio_writeprotect.mode
+should be set to
+.BR UFFDIO_WRITEPROTECT_MODE_WP .
+.PP
+When a write-protect event happens, the userspace will receive a page fault
+message whose
+.I uffd_msg.pagefault.flags
+will be with
+.BR UFFD_PAGEFAULT_FLAG_WP
+flag set.  Note: since only writes can trigger such kind of fault,
+write-protect messages will always be with
+.BR UFFD_PAGEFAULT_FLAG_WRITE
+bit set too along with
+.BR UFFD_PAGEFAULT_FLAG_WP .
+.PP
+Currently, write-protect mode only supports private anonymous memory.
 .SS Reading from the userfaultfd structure
 Each
 .BR read (2)
@@ -363,8 +443,12 @@ flag (see
 .BR ioctl_userfaultfd (2))
 and this flag is set, this a write fault;
 otherwise it is a read fault.
-.\"
-.\" UFFD_PAGEFAULT_FLAG_WP is not yet supported.
+.TP
+.B UFFD_PAGEFAULT_FLAG_WP
+If the address is in a range that was registered with the
+.B UFFDIO_REGISTER_MODE_WP
+flag, when this bit is set it means it's a write-protect fault.  Otherwise it's
+a page missing fault.
 .RE
 .TP
 .I pagefault.feat.pid
Signature · 1 line
2.26.2
Signature · 2 lines
Sincerely yours,
Mike.
author
#9 lore raw

Re: [PATCH 2/4] userfaultfd.2: Add write-protect mode

Quoted 41 lines · Mike Rapoport
Quoted 27 lines · Peter Xu

Write-protect mode is supported starting from Linux 5.7.

---
 man2/userfaultfd.2 | 88 ++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 86 insertions(+), 2 deletions(-)
man2/userfaultfd.2 +12 -0
index 2d14effc6..8e1602d62 100644
@@ -78,6 +78,28 @@ all memory ranges that were registered with the object are unregistered
 and unread events are flushed.
 .\"
 .PP
+Currently, userfaultfd supports two modes of registration:
+.TP
+.BR UFFDIO_REGISTER_MODE_MISSING
+When registered with
+.BR UFFDIO_REGISTER_MODE_MISSING
+mode, the userspace will receive a page fault message when a missing page is
+accessed.  The faulted thread will be stopped from execution until the page
+fault is resolved from the userspace by either an
+.BR UFFDIO_COPY
+or an
+.BR UFFDIO_ZEROPAGE
+ioctl.

man2/userfaultfd.2 all memory ranges that were registered with the object are unregistered · line 102

Quoted 10 lines
man2/userfaultfd.2 +10 -0
+.TP
+.BR UFFDIO_REGISTER_MODE_WP
+When registered with
+.BR UFFDIO_REGISTER_MODE_WP
+mode, the userspace will receive a page fault message when a write-protected
+page is written.  The faulted thread will be stopped from execution until the
+userspace un-write-protect the page using an
+.BR UFFDIO_WRITEPROTECT
+ioctl.
+.PP

I'd add a sentence about combining the modes together. Something like

Quoted 1 line

"Both modes can be enabled together for the same memory range"

I mentioned it below [1]. However I agree it's indeed making more sense to mention it when listing the modes, especially knowing that the 3rd minor mode is coming. I think I'll keep both, assuming a bit more verbose is still acceptable in man pages, but changed to:

"Multiple modes can be enabled at the same time for the same memory range."

Quoted 21 lines
Quoted 21 lines
man2/userfaultfd.2 +10 -0
 Since Linux 4.14, userfaultfd page fault message can selectively embed fault
 thread ID information into the fault message.  One needs to enable this feature
 explicitly using the
@@ -143,6 +165,16 @@ single threaded non-cooperative userfaultfd manager implementations.
 .\" and limitations remaining in 4.11
 .\" Maybe it's worth adding a dedicated sub-section...
 .\"
+.PP
+Starting from Linux 5.7, userfaultfd is able to do synchronous page dirty
+tracking using the new write-protection register mode.  One should check
+against the feature bit
+.B UFFD_FEATURE_PAGEFAULT_FLAG_WP
+before using this feature.  Similar to the original userfaultfd missing mode,
+the write-protect mode will generate an userfaultfd message when the protected
+page is written.  The user needs to resolve the page fault by unprotecting the
+faulted page and kick the faulted thread to continue.  For more information,
+please read the "Userfaultfd write-protect mode" section below.
 .SS Userfaultfd operation
 After the userfaultfd object is created with
 .BR userfaultfd (),
Quoted 8 lines

man2/userfaultfd.2 userfaultfd can be used only with anonymous private memory mappings. · line 254

Quoted 6 lines
man2/userfaultfd.2 +2 -0
@@ -218,6 +250,54 @@ userfaultfd can be used only with anonymous private memory mappings.
 Since Linux 4.11,
 userfaultfd can be also used with hugetlbfs and shared memory mappings.
 .\"
+.SS Userfaultfd write-protect mode
+Since Linux 5.7, userfaultfd started to support write-protect mode.  The user

Maybe s/started to support/supports/

Sure.

Quoted 15 lines
Quoted 5 lines
man2/userfaultfd.2 +4 -0
+needs to first check availability of this feature using
+.BR UFFDIO_API
+ioctl against the feature bit
+.BR UFFD_FEATURE_PAGEFAULT_FLAG_WP .

man2/userfaultfd.2 userfaultfd can be used only with anonymous private memory mappings. · line 268

Quoted 10 lines
man2/userfaultfd.2 +10 -0
+.PP
+To register with userfaultfd write-protect mode, the user needs to send the
+.BR UFFDIO_REGISTER
+ioctl with mode
+.BR UFFDIO_REGISTER_MODE_WP
+set.  Note that it's legal to monitor the same memory range with multiple
+modes.  For example, the user can do
+.BR UFFDIO_REGISTER
+with the mode set to
+.BR UFFDIO_REGISTER_MODE_MISSING\ |\ UFFDIO_REGISTER_MODE_WP.

[1]

Thanks,

Signature · 1 line
Peter Xu
author
#10 lore raw

PATCH 3/4 ioctl_userfaultfd.2: Add UFFD_FEATURE_THREAD_ID docs

In reply to Peter Xu

UFFD_FEATURE_THREAD_ID is supported in Linux 4.14.

---
 man2/ioctl_userfaultfd.2 | 4 ++++
 1 file changed, 4 insertions(+)
man2/ioctl_userfaultfd.2 +4 -0
index 47ae5f473..1965d1932 100644
@@ -208,6 +208,10 @@ signal will be sent to the faulting process.
 Applications using this
 feature will not require the use of a userfaultfd monitor for processing
 memory accesses to the regions registered with userfaultfd.
+.BR UFFD_FEATURE_THREAD_ID " (since Linux 4.14)"
+If this feature bit is set,
+.I uffd_msg.pagefault.feat.ptid
+Will be set
 .PP
 The returned
 .I ioctls
Signature · 1 line
2.26.2
author
#11 lore raw

PATCH 4/4 ioctl_userfaultfd.2: Add write-protect mode docs

In reply to Peter Xu

Userfaultfd write-protect mode is supported starting from Linux 5.7.

---
 man2/ioctl_userfaultfd.2 | 74 +++++++++++++++++++++++++++++++++++++---
 1 file changed, 70 insertions(+), 4 deletions(-)
man2/ioctl_userfaultfd.2 +70 -4
index 1965d1932..3feb888a8 100644
@@ -208,10 +208,11 @@ signal will be sent to the faulting process.
 Applications using this
 feature will not require the use of a userfaultfd monitor for processing
 memory accesses to the regions registered with userfaultfd.
+.TP
 .BR UFFD_FEATURE_THREAD_ID " (since Linux 4.14)"
 If this feature bit is set,
 .I uffd_msg.pagefault.feat.ptid
-Will be set
+will be set to the faulted thread ID for each page fault message.
 .PP
 The returned
 .I ioctls
@@ -233,6 +234,11 @@ operation is supported.
 The
 .B UFFDIO_UNREGISTER
 operation is supported.
+.TP
+.B 1 << _UFFDIO_WRITEPROTECT
+The
+.B UFFDIO_WRITEPROTECT
+operation is supported.
 .PP
 This
 .BR ioctl (2)
@@ -321,9 +327,6 @@ Track page faults on missing pages.
 .B UFFDIO_REGISTER_MODE_WP
 Track page faults on write-protected pages.
 .PP
-Currently, the only supported mode is
-.BR UFFDIO_REGISTER_MODE_MISSING .
-.PP
 If the operation is successful, the kernel modifies the
 .I ioctls
 bit-mask field to indicate which
@@ -653,6 +656,69 @@ field of the
 structure was not a multiple of the system page size; or
 .I len
 was zero; or the specified range was otherwise invalid.
+.SS UFFDIO_WRITEPROTECT
+(Since Linux 5.7) Do write-protect or write-unprotect for an userfaultfd
+registered memory range with mode
+.BR UFFDIO_REGISTER_MODE_WP .
+.PP
+The
+.I argp
+argument is a pointer to a
+.I uffdio_range
+structure as shown below:
+.PP
+.in +4n
+.EX
+struct uffdio_writeprotect {
+    struct uffdio_range range;  /* Range to change write permission */
+    __u64 mode;                 /* Mode to change write permission */
+};
+.EE
+.in
+There're two modes that are supported in this structure:
+.TP
+.B UFFDIO_WRITEPROTECT_MODE_WP
+When this mode bit is set, the ioctl will be a write-protect operation upon the
+memory range specified by
+.IR range .
+Otherwise it'll be a write-unprotect operation upon the specified range.
+.TP
+.B UFFDIO_WRITEPROTECT_MODE_DONTWAKE
+Do not wake up the thread that waits for page-fault resolution after the
+operation.  This could only be specified if
+.B UFFDIO_WRITEPROTECT_MODE_WP
+is not specified (in a resolving stage, not protecting stage).
+.PP
+This
+.BR ioctl (2)
+operation returns 0 on success.
+On error, \-1 is returned and
+.I errno
+is set to indicate the error.
+Possible errors include:
+.TP
+.B EINVAL
+The
+.I start
+or the
+.I len
+field of the
+.I ufdio_range
+structure was not a multiple of the system page size; or
+.I len
+was zero; or the specified range was otherwise invalid.
+.TP
+.B EAGAIN
+The process was interrupted and need to retry.
+.TP
+.B ENOENT
+The range specified in
+.I range
+is not valid.  E.g., the virtual address does not exist, or not registered with
+userfaultfd write-protect mode.
+.TP
+.B EFAULT
+Encountered a generic fault during processing.
 .SH RETURN VALUE
 See descriptions of the individual operations, above.
 .SH ERRORS
Signature · 1 line
2.26.2
author
#12 lore raw

Re: [PATCH 4/4] ioctl_userfaultfd.2: Add write-protect mode docs

Quoted 12 lines · Peter Xu

Userfaultfd write-protect mode is supported starting from Linux 5.7.

---
 man2/ioctl_userfaultfd.2 | 74 +++++++++++++++++++++++++++++++++++++---
 1 file changed, 70 insertions(+), 4 deletions(-)
man2/ioctl_userfaultfd.2 +0 -0
index 1965d1932..3feb888a8 100644

man2/ioctl_userfaultfd.2 signal will be sent to the faulting process. · line 215

Quoted 9 lines
man2/ioctl_userfaultfd.2 +2 -1
@@ -208,10 +208,11 @@ signal will be sent to the faulting process.
 Applications using this
 feature will not require the use of a userfaultfd monitor for processing
 memory accesses to the regions registered with userfaultfd.
+.TP
 .BR UFFD_FEATURE_THREAD_ID " (since Linux 4.14)"
 If this feature bit is set,
 .I uffd_msg.pagefault.feat.ptid
-Will be set
+will be set to the faulted thread ID for each page fault message.

These two changes should belong to the previous patch - will get them fixed too.

Signature · 1 line
Peter Xu