clone.2: tfix
PATCH 1/2 clone.2: tfix
added two missing parentheses
--- man2/clone.2 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
man2/clone.2 +2 -2
@@ -1349,13 +1349,13 @@ For example, on aarch64, .I stack must be a multiple of 16. .TP -.BR EINVAL " (" clone3 "() only" +.BR EINVAL " (" clone3 "() only)" .B CLONE_DETACHED was specified in the .I flags mask. .TP -.BR EINVAL " (" clone "() only" +.BR EINVAL " (" clone "() only)" .B CLONE_PIDFD was specified together with .B CLONE_DETACHED
base-commit: 445fc03eebeee9b0e9ccaa4e2361f91f30749e8e
Signature · 1 line
2.23.0
PATCH 2/2 clone.2: added clone3() set_tid information
--- man2/clone.2 | 90 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 90 insertions(+)
man2/clone.2 +90 -0
@@ -195,6 +195,8 @@ struct clone_args { u64 stack; /* Pointer to lowest byte of stack */ u64 stack_size; /* Size of stack */ u64 tls; /* Location of new TLS */ + u64 set_tid; /* Pointer to a \fIpid_t\fP array */ + u64 set_tid_size; /* Number of elements in \fIset_tid\fP */ }; .EE .in
@@ -262,6 +264,8 @@ flags & 0xff exit_signal stack stack \fP---\fP stack_size tls tls See CLONE_SETTLS +\fP---\fP set_tid See below for details +\fP---\fP set_tid_size .TE .RE .\"
@@ -285,6 +289,74 @@ options when waiting for the child with If no signal (i.e., zero) is specified, then the parent process is not signaled when the child terminates. .\" +.SS The set_tid array +.PP +The +.I set_tid +array is used to select a certain PID for the process to be created by +.BR clone3 (). +If the PID of the newly created process should only be set for the current +PID namespace or in the newly created PID namespace (if +.I flags +contains +.BR CLONE_NEWPID ) +then the first element in the +.I set_tid +array has to be the desired PID and +.I set_tid_size +needs to be 1. +.PP +If the PID of the newly created process should have a certain value in +multiple PID namespaces the +.I set_tid +array can have multiple entries. The first entry defines the PID in the most +nested PID namespace and all following entries contain the PID of the +corresponding parent PID namespace. The number of PID namespaces in which a PID +should be set is defined by +.I set_tid_size +which cannot be larger than the number of currently nested PID namespaces. +.PP +To create a process with the following PIDs: +.RS +.TS +lb lb +l l . +PID NS level Requested PID +0 (host) 31496 +1 42 +2 7 +.TE +.RE +.PP +The +.I set_tid +array would need to be filled with: +.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid[2] = 31496; + set_tid_size = 3; +.EE +.PP +If only the PID of the two innermost PID namespaces +should be defined it needs to be set like this: +.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid_size = 2; +.EE +.PP +The PID in the PID namespaces outside the two innermost PID namespaces +is then selected the same way as any other PID is selected. +.PP +Only a privileged process +.RB ( CAP_SYS_ADMIN ) +can set +.I set_tid +to select a PID for the process to be created. +.\" .SS The flags mask .PP Both
@@ -1379,6 +1451,16 @@ in the .I flags mask. .TP +.BR EINVAL " (" clone3 "() only)" +.I set_tid_size +larger than current number of nested PID namespaces or maximum number of +nested PID namespaces was specified. +.TP +.BR EINVAL " (" clone3 "() only)" +If one of the PIDs specified in +.I set_tid +was an invalid PID. +.TP .B ENOMEM Cannot allocate sufficient memory to allocate a task structure for the child, or to copy those parts of the caller's context that need to be
@@ -1450,6 +1532,14 @@ mask and the caller is in a chroot environment (i.e., the caller's root directory does not match the root directory of the mount namespace in which it resides). .TP +.BR EPERM " (" clone3 "() only)" +If +.I set_tid +with +.I set_tid_size +larger than 0 was specified by an unprivileged process (process without +\fBCAP_SYS_ADMIN\fP). +.TP .BR ERESTARTNOINTR " (since Linux 2.6.17)" .\" commit 4a2c7a7837da1b91468e50426066d988050e4d56 System call was interrupted by a signal and will be restarted.
Signature · 1 line
2.23.0
Re: [PATCH 2/2] clone.2: added clone3() set_tid information
Quoted 47 lines · Adrian Reber
Signed-off-by Adrian Reber <[email protected]>--- man2/clone.2 | 90 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 90 insertions(+)man2/clone.2 +20
-0@@ -195,6 +195,8 @@ struct clone_args { u64 stack; /* Pointer to lowest byte of stack */ u64 stack_size; /* Size of stack */ u64 tls; /* Location of new TLS */ + u64 set_tid; /* Pointer to a \fIpid_t\fP array */ + u64 set_tid_size; /* Number of elements in \fIset_tid\fP */ }; .EE .in@@ -262,6 +264,8 @@ flags & 0xff exit_signal stack stack \fP---\fP stack_size tls tls See CLONE_SETTLS +\fP---\fP set_tid See below for details +\fP---\fP set_tid_size .TE .RE .\"@@ -285,6 +289,74 @@ options when waiting for the child with If no signal (i.e., zero) is specified, then the parent process is not signaled when the child terminates. .\" +.SS The set_tid array +.PP +The +.I set_tid +array is used to select a certain PID for the process to be created by +.BR clone3 (). +If the PID of the newly created process should only be set for the current +PID namespace or in the newly created PID namespace (if +.I flags +contains +.BR CLONE_NEWPID ) +then the first element in the +.I set_tid +array has to be the desired PID and +.I set_tid_size +needs to be 1.
man2/clone.2 options when waiting for the child with · line 317
Quoted 10 lines
man2/clone.2 +10
-0+.PP +If the PID of the newly created process should have a certain value in +multiple PID namespaces the +.I set_tid +array can have multiple entries. The first entry defines the PID in the most +nested PID namespace and all following entries contain the PID of the +corresponding parent PID namespace. The number of PID namespaces in which a PID +should be set is defined by +.I set_tid_size +which cannot be larger than the number of currently nested PID namespaces.
"It's upper cap is the kernel-enforced general nesting limit." or sm like that
man2/clone.2 options when waiting for the child with · line 319
Quoted 2 lines
man2/clone.2 +2
-0+.PP +To create a process with the following PIDs:
in a pid namespace hierarchy
man2/clone.2 options when waiting for the child with · line 330
Quoted 11 lines
man2/clone.2 +11
-0+.RS +.TS +lb lb +l l . +PID NS level Requested PID +0 (host) 31496 +1 42 +2 7 +.TE +.RE +.PP
Set the array to:
man2/clone.2 options when waiting for the child with · line 342
Quoted 9 lines
man2/clone.2 +9
-0+.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid[2] = 31496; + set_tid_size = 3; +.EE +.PP +If only the PID of the two innermost PID namespaces
need to be specified set the array to:
man2/clone.2 options when waiting for the child with · line 352
Quoted 9 lines
man2/clone.2 +9
-0+.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid_size = 2; +.EE +.PP +The PID in the PID namespaces outside the two innermost PID namespaces +is then selected the same way as any other PID is selected.
s/is then/will be/
man2/clone.2 options when waiting for the child with · line 355
Quoted 3 lines
man2/clone.2 +3
-0+.PP +Only a privileged process +.RB ( CAP_SYS_ADMIN )
Maybe more specific:
The set_tid feature requires CAP_SYS_ADMIN in all owning user namespaces of the target pid namespaces.
Quoted 25 lines
man2/clone.2 +14
-0+can set +.I set_tid +to select a PID for the process to be created. +.\" .SS The flags mask .PP Both@@ -1379,6 +1451,16 @@ in the .I flags mask. .TP +.BR EINVAL " (" clone3 "() only)" +.I set_tid_size +larger than current number of nested PID namespaces or maximum number of +nested PID namespaces was specified. +.TP +.BR EINVAL " (" clone3 "() only)" +If one of the PIDs specified in +.I set_tid +was an invalid PID. +.TP .B ENOMEM Cannot allocate sufficient memory to allocate a task structure for the child, or to copy those parts of the caller's context that need to be
man2/clone.2 mask and the caller is in a chroot environment · line 1541
Quoted 10 lines
man2/clone.2 +7
-0@@ -1450,6 +1532,14 @@ mask and the caller is in a chroot environment (i.e., the caller's root directory does not match the root directory of the mount namespace in which it resides). .TP +.BR EPERM " (" clone3 "() only)" +If +.I set_tid +with +.I set_tid_size +larger than 0 was specified by an unprivileged process (process without +\fBCAP_SYS_ADMIN\fP).
without CAP_SYS_ADMIN in _one_ of the owning user namespaces of the target pid namespace
maybe?
Where's EEXIST?
Christian
Re: [PATCH 2/2] clone.2: added clone3() set_tid information
Quoted 53 lines · Christian Brauner
Quoted 52 lines · Adrian Reber
Signed-off-by Adrian Reber <[email protected]>--- man2/clone.2 | 90 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 90 insertions(+)man2/clone.2 +25
-0@@ -195,6 +195,8 @@ struct clone_args { u64 stack; /* Pointer to lowest byte of stack */ u64 stack_size; /* Size of stack */ u64 tls; /* Location of new TLS */ + u64 set_tid; /* Pointer to a \fIpid_t\fP array */ + u64 set_tid_size; /* Number of elements in \fIset_tid\fP */ }; .EE .in@@ -262,6 +264,8 @@ flags & 0xff exit_signal stack stack \fP---\fP stack_size tls tls See CLONE_SETTLS +\fP---\fP set_tid See below for details +\fP---\fP set_tid_size .TE .RE .\"@@ -285,6 +289,74 @@ options when waiting for the child with If no signal (i.e., zero) is specified, then the parent process is not signaled when the child terminates. .\" +.SS The set_tid array +.PP +The +.I set_tid +array is used to select a certain PID for the process to be created by +.BR clone3 (). +If the PID of the newly created process should only be set for the current +PID namespace or in the newly created PID namespace (if +.I flags +contains +.BR CLONE_NEWPID ) +then the first element in the +.I set_tid +array has to be the desired PID and +.I set_tid_size +needs to be 1. +.PP +If the PID of the newly created process should have a certain value in +multiple PID namespaces the +.I set_tid +array can have multiple entries. The first entry defines the PID in the most
Quoted 8 lines
man2/clone.2 options when waiting for the child with · line 317
Quoted 5 lines
man2/clone.2 +5
-0+nested PID namespace and all following entries contain the PID of the +corresponding parent PID namespace. The number of PID namespaces in which a PID +should be set is defined by +.I set_tid_size +which cannot be larger than the number of currently nested PID namespaces."It's upper cap is the kernel-enforced general nesting limit." or sm like that
Is that an addition to my sentence or a replacement. I think at this point it is more important to point out that it cannot be larger than the number of currently nested PID namespaces. Later (at EPERM) I am also mentioning that it cannot be larger than the maximum number of nested PID namespaces. The code does indeed check if set_tid_size is larger than the maximum number of possible nested PID namespaces for the user, I think, when calling clone3(), it is more relevant that set_tid_size is not larger than the number of currently nested PID namespaces. The maximum number of possible nested PID namespaces is more likely enforced during unshare() or CLONE_NEWPID (which could be happening at the same point in time as set_tid_size larger than maximum number of nested PID namespace).
This definitely feels like too much discussion for a single sentence ;)
I can add a sentence about the maximum number of nested PID namespaces here in addition to the one at EPERM.
I do not think it is relevant for the user at this point in time.
Adrian
Re: [PATCH 2/2] clone.2: added clone3() set_tid information
Hi Adrian,
Thank you for the patch. I have some review comments below, and also further comments in my reply to Christian's notes.
Quoted 76 lines · Adrian Reber
Quoted 53 lines · Christian Brauner
Quoted 52 lines · Adrian Reber
Signed-off-by Adrian Reber <[email protected]>--- man2/clone.2 | 90 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 90 insertions(+)man2/clone.2 +25
-0@@ -195,6 +195,8 @@ struct clone_args { u64 stack; /* Pointer to lowest byte of stack */ u64 stack_size; /* Size of stack */ u64 tls; /* Location of new TLS */ + u64 set_tid; /* Pointer to a \fIpid_t\fP array */ + u64 set_tid_size; /* Number of elements in \fIset_tid\fP */ }; .EE .in@@ -262,6 +264,8 @@ flags & 0xff exit_signal stack stack \fP---\fP stack_size tls tls See CLONE_SETTLS +\fP---\fP set_tid See below for details +\fP---\fP set_tid_size .TE .RE .\"@@ -285,6 +289,74 @@ options when waiting for the child with If no signal (i.e., zero) is specified, then the parent process is not signaled when the child terminates. .\" +.SS The set_tid array +.PP +The +.I set_tid +array is used to select a certain PID for the process to be created by +.BR clone3 (). +If the PID of the newly created process should only be set for the current +PID namespace or in the newly created PID namespace (if +.I flags +contains +.BR CLONE_NEWPID ) +then the first element in the +.I set_tid +array has to be the desired PID and +.I set_tid_size +needs to be 1. +.PP +If the PID of the newly created process should have a certain value in +multiple PID namespaces the +.I set_tid +array can have multiple entries. The first entry defines the PID in the mostQuoted 8 lines
man2/clone.2 options when waiting for the child with · line 317
Quoted 5 lines
man2/clone.2 +5
-0+nested PID namespace and all following entries contain the PID of the +corresponding parent PID namespace. The number of PID namespaces in which a PID +should be set is defined by +.I set_tid_size +which cannot be larger than the number of currently nested PID namespaces."It's upper cap is the kernel-enforced general nesting limit." or sm like that
Is that an addition to my sentence or a replacement. I think at this point it is more important to point out that it cannot be larger than the number of currently nested PID namespaces. Later (at EPERM) I am also mentioning that it cannot be larger than the maximum number of nested PID namespaces. The code does indeed check if
set_tid_sizeis larger than the maximum number of possible nested PID namespaces for the user, I think, when callingclone3(), it is more relevant thatset_tid_sizeis not larger than the number of currently nested PID namespaces. The maximum number of possible nested PID namespaces is more likely enforced duringunshare()orCLONE_NEWPID(which could be happening at the same point in time asset_tid_sizelarger than maximum number of nested PID namespace).
Quoted 5 lines
This definitely feels like too much discussion for a single sentence ;)
I can add a sentence about the maximum number of nested PID namespaces here in addition to the one at EPERM.
I do not think it is relevant for the user at this point in time.
I'm inclined to agree with you. The maximum nesting depth is a general detail of PID NSs and is already covered in pid_namespaces((7). I think there's no need to repeat or allude to that detial here the detail here.
Thanks,
Michael
Re: [PATCH 2/2] clone.2: added clone3() set_tid information
In reply to Christian BraunerHello Christian, and Adrian,
Christian, thanks for the review!
Quoted 53 lines · Christian Brauner
Quoted 52 lines · Adrian Reber
Signed-off-by Adrian Reber <[email protected]>--- man2/clone.2 | 90 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 90 insertions(+)man2/clone.2 +25
-0@@ -195,6 +195,8 @@ struct clone_args { u64 stack; /* Pointer to lowest byte of stack */ u64 stack_size; /* Size of stack */ u64 tls; /* Location of new TLS */ + u64 set_tid; /* Pointer to a \fIpid_t\fP array */ + u64 set_tid_size; /* Number of elements in \fIset_tid\fP */ }; .EE .in@@ -262,6 +264,8 @@ flags & 0xff exit_signal stack stack \fP---\fP stack_size tls tls See CLONE_SETTLS +\fP---\fP set_tid See below for details +\fP---\fP set_tid_size .TE .RE .\"@@ -285,6 +289,74 @@ options when waiting for the child with If no signal (i.e., zero) is specified, then the parent process is not signaled when the child terminates. .\" +.SS The set_tid array +.PP +The +.I set_tid +array is used to select a certain PID for the process to be created by +.BR clone3 (). +If the PID of the newly created process should only be set for the current +PID namespace or in the newly created PID namespace (if +.I flags +contains +.BR CLONE_NEWPID ) +then the first element in the +.I set_tid +array has to be the desired PID and +.I set_tid_size +needs to be 1. +.PP +If the PID of the newly created process should have a certain value in +multiple PID namespaces the +.I set_tid +array can have multiple entries. The first entry defines the PID in the most
Quoted 8 lines
man2/clone.2 options when waiting for the child with · line 317
Quoted 5 lines
man2/clone.2 +5
-0+nested PID namespace and all following entries contain the PID of the +corresponding parent PID namespace. The number of PID namespaces in which a PID +should be set is defined by +.I set_tid_size +which cannot be larger than the number of currently nested PID namespaces."It's upper cap is the kernel-enforced general nesting limit." or sm like that
See my comments in a separate mail where Adrian replied to the above.
Quoted 4 lines
man2/clone.2 options when waiting for the child with · line 319
Quoted 2 lines
man2/clone.2 +2
-0+.PP +To create a process with the following PIDs:in a pid namespace hierarchy
(okay)
Quoted 12 lines
man2/clone.2 options when waiting for the child with · line 330
Quoted 11 lines
man2/clone.2 +11
-0+.RS +.TS +lb lb +l l . +PID NS level Requested PID +0 (host) 31496 +1 42 +2 7 +.TE +.RE +.PP
Christian, you trim away the piece:
[[
+The
+.I set_tid
+array would need to be filled with:
]]
that you suggest replacing with:
Quoted 1 line
Set the array to:
I think the suggested change is fine, but when you trim the text in this way (as you do a few times below, it makes it overly difficult to review your suggestions, since I must separately refer back to Adrian's original mail to see what is being replaced (which is not comfortable on my 13" laptop 11km above Belgium :-)).
Quoted 9 lines
man2/clone.2 options when waiting for the child with · line 342
Quoted 9 lines
man2/clone.2 +9
-0+.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid[2] = 31496; + set_tid_size = 3; +.EE +.PP +If only the PID of the two innermost PID namespaces
s/PID of/PIDs in/
Quoted 2 lines
need to be specified set the array to:
(okay; but see my comment about trimming above.)
Quoted 11 lines
man2/clone.2 options when waiting for the child with · line 352
Quoted 9 lines
man2/clone.2 +9
-0+.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid_size = 2; +.EE +.PP +The PID in the PID namespaces outside the two innermost PID namespaces +is then selected the same way as any other PID is selected.s/is then/will be/
(okay)
Quoted 8 lines
man2/clone.2 options when waiting for the child with · line 355
Quoted 3 lines
man2/clone.2 +3
-0+.PP +Only a privileged process +.RB ( CAP_SYS_ADMIN )Maybe more specific:
The
set_tidfeature requiresCAP_SYS_ADMINin all owning user namespaces of the target pid namespaces.
Excellent addition.
Quoted 7 lines
Quoted 7 lines
man2/clone.2 +4
-0+can set +.I set_tid +to select a PID for the process to be created. +.\" .SS The flags mask .PP Both
Quoted 8 lines
man2/clone.2 in the · line 1457
Quoted 8 lines
man2/clone.2 +4
-0@@ -1379,6 +1451,16 @@ in the .I flags mask. .TP +.BR EINVAL " (" clone3 "() only)" +.I set_tid_size +larger than current number of nested PID namespaces or maximum number of +nested PID namespaces was specified.
The last piece of that sentence feels redundant to me, and it could/should be removed, I think. The current number of nested PID namespaces can't exceed the maximum number of nested namespaces, so mentioning the latter as a limit seems unnecessary (and even a little confusing, since it leaves the reader wondering whether "current" could exceed "maximum".)
Or have I misunderstood what you intend to convey, Adrian?
Quoted 24 lines
Quoted 10 lines
man2/clone.2 +6
-0+.TP +.BR EINVAL " (" clone3 "() only)" +If one of the PIDs specified in +.I set_tid +was an invalid PID. +.TP .B ENOMEM Cannot allocate sufficient memory to allocate a task structure for the child, or to copy those parts of the caller's context that need to beman2/clone.2 mask and the caller is in a chroot environment · line 1541
Quoted 10 lines
man2/clone.2 +7
-0@@ -1450,6 +1532,14 @@ mask and the caller is in a chroot environment (i.e., the caller's root directory does not match the root directory of the mount namespace in which it resides). .TP +.BR EPERM " (" clone3 "() only)" +If +.I set_tid +with +.I set_tid_size +larger than 0 was specified by an unprivileged process (process without +\fBCAP_SYS_ADMIN\fP).without
CAP_SYS_ADMINin _one_ of the owning user namespaces of the target pid namespace
Quoted 1 line
maybe?
Adrian, I think there's no need to mention 'set_tid'. That's implied by 'set_tid_size' being greater than 0.
How about:
[[
.BR EPERM " (" clone3 "() only)"
.I set_tid_size
was greater than zero, and the caller lacks the
.B CAP_SYS_ADMIN
capability in one or more of the user namespaces that own the corresponding PID namespaces.
]]
?
Quoted 1 line
Where's EEXIST?
Indeed!
I speculate (without having looked at the kernel code):
[[
.BR EPERM " (" clone3 "() only)"
One or more of the PIDs specified in
.I set_tid
already exists in the corresponding namespace.
]]
Thanks,
Michael
Re: [PATCH 2/2] clone.2: added clone3() set_tid information
Christian and Michael,
thanks for the reviews. I tried to apply all suggested changes in v2 which I will send out in a few minutes.
Adrian
Re: [PATCH 2/2] clone.2: added clone3() set_tid information
In reply to Adrian ReberHello Adrian,
Quoted 28 lines · Adrian Reber
Signed-off-by Adrian Reber <[email protected]>--- man2/clone.2 | 90 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 90 insertions(+)man2/clone.2 +4
-0@@ -195,6 +195,8 @@ struct clone_args { u64 stack; /* Pointer to lowest byte of stack */ u64 stack_size; /* Size of stack */ u64 tls; /* Location of new TLS */ + u64 set_tid; /* Pointer to a \fIpid_t\fP array */ + u64 set_tid_size; /* Number of elements in \fIset_tid\fP */ }; .EE .in@@ -262,6 +264,8 @@ flags & 0xff exit_signal stack stack \fP---\fP stack_size tls tls See CLONE_SETTLS +\fP---\fP set_tid See below for details +\fP---\fP set_tid_size .TE .RE .\"
man2/clone.2 options when waiting for the child with · line 296
Quoted 8 lines
man2/clone.2 +5
-0@@ -285,6 +289,74 @@ options when waiting for the child with If no signal (i.e., zero) is specified, then the parent process is not signaled when the child terminates. .\" +.SS The set_tid array +.PP +The +.I set_tid +array is used to select a certain PID for the process to be created by
s/is used/may be used/
Because it's not required to use this array, right? I mean, the default is that the kernel chooses the PIDs. Perhaps this needs to be more clearly stated at the start of this subsection.
How about:
[[
By default, the kernel chooses the next sequential PID for the new process in each of the PID namespaces where it is present. When creating a process with
.BR clone3 (),
the
.I set_tid
array can be used to select specific PIDs for the process in some or all of the PID namespaces where it is present.
]]
?
Quoted 6 lines
man2/clone.2 +6
-0+.BR clone3 (). +If the PID of the newly created process should only be set for the current +PID namespace or in the newly created PID namespace (if +.I flags +contains +.BR CLONE_NEWPID )
man2/clone.2 options when waiting for the child with · line 312
Quoted 10 lines
man2/clone.2 +10
-0+then the first element in the +.I set_tid +array has to be the desired PID and +.I set_tid_size +needs to be 1. +.PP +If the PID of the newly created process should have a certain value in +multiple PID namespaces the +.I set_tid +array can have multiple entries. The first entry defines the PID in the most
most *deeply* nested
Quoted 72 lines
man2/clone.2 +58
-0+nested PID namespace and all following entries contain the PID of the +corresponding parent PID namespace. The number of PID namespaces in which a PID +should be set is defined by +.I set_tid_size +which cannot be larger than the number of currently nested PID namespaces. +.PP +To create a process with the following PIDs: +.RS +.TS +lb lb +l l . +PID NS level Requested PID +0 (host) 31496 +1 42 +2 7 +.TE +.RE +.PP +The +.I set_tid +array would need to be filled with: +.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid[2] = 31496; + set_tid_size = 3; +.EE +.PP +If only the PID of the two innermost PID namespaces +should be defined it needs to be set like this: +.PP +.EX + set_tid[0] = 7; + set_tid[1] = 42; + set_tid_size = 2; +.EE +.PP +The PID in the PID namespaces outside the two innermost PID namespaces +is then selected the same way as any other PID is selected. +.PP +Only a privileged process +.RB ( CAP_SYS_ADMIN ) +can set +.I set_tid +to select a PID for the process to be created. +.\" .SS The flags mask .PP Both@@ -1379,6 +1451,16 @@ in the .I flags mask. .TP +.BR EINVAL " (" clone3 "() only)" +.I set_tid_size +larger than current number of nested PID namespaces or maximum number of +nested PID namespaces was specified. +.TP +.BR EINVAL " (" clone3 "() only)" +If one of the PIDs specified in +.I set_tid +was an invalid PID. +.TP .B ENOMEM Cannot allocate sufficient memory to allocate a task structure for the child, or to copy those parts of the caller's context that need to be@@ -1450,6 +1532,14 @@ mask and the caller is in a chroot environment (i.e., the caller's root directory does not match the root directory of the mount namespace in which it resides). .TP +.BR EPERM " (" clone3 "() only)"
man2/clone.2 mask and the caller is in a chroot environment · line 1545
Quoted 10 lines
man2/clone.2 +7
-0+If +.I set_tid +with +.I set_tid_size +larger than 0 was specified by an unprivileged process (process without +\fBCAP_SYS_ADMIN\fP). +.TP .BR ERESTARTNOINTR " (since Linux 2.6.17)" .\" commit 4a2c7a7837da1b91468e50426066d988050e4d56 System call was interrupted by a signal and will be restarted.
Thanks,
Michael
Quoted 3 lines · Adrian Reber
added two missing parentheses
Signed-off-by Adrian Reber <[email protected]>
Seems reasonable to me.
Hi Adrian,
Quoted 28 lines · Adrian Reber
added two missing parentheses
Signed-off-by Adrian Reber <[email protected]>--- man2/clone.2 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)man2/clone.2 +2
-2@@ -1349,13 +1349,13 @@ For example, on aarch64, .I stack must be a multiple of 16. .TP -.BR EINVAL " (" clone3 "() only" +.BR EINVAL " (" clone3 "() only)" .B CLONE_DETACHED was specified in the .I flags mask. .TP -.BR EINVAL " (" clone "() only" +.BR EINVAL " (" clone "() only)" .B CLONE_PIDFD was specified together with .B CLONE_DETACHED
Quoted 1 line
base-commit:
445fc03eebeee9b0e9ccaa4e2361f91f30749e8e
Thanks. Applied.
Cheers,
Michael