ioctl()-based API to query VMAs from /proc/<pid>/maps
Subsystems (7)
Andrii Nakryiko
maintainerv5 cover ioctl()-based API to query VMAs from /proc/<pid>/maps
Implement binary ioctl()-based interface to /proc/<pid>/maps file to allow applications to query VMA information more efficiently than reading *all* VMAs nonselectively through text-based interface of /proc/<pid>/maps file.
Patch #2 goes into a lot of details and background on some common patterns of using /proc/<pid>/maps in the area of performance profiling and subsequent symbolization of captured stack traces. As mentioned in that patch, patterns of VMA querying can differ depending on specific use case, but can generally be grouped into two main categories: the need to query a small subset of VMAs covering a given batch of addresses, or reading/storing/caching all (typically, executable) VMAs upfront for later processing.
The new PROCMAP_QUERY ioctl() API added in this patch set was motivated by the former pattern of usage. Earlier revisions had a patch adding a tool that faithfully reproduces an efficient VMA matching pass of a symbolizer, collecting a subset of covering VMAs for a given set of addresses as efficiently as possible. This tool served both as a testing ground, as well as a benchmarking tool. It implements everything both for currently existing text-based /proc/<pid>/maps interface, as well as for newly-added PROCMAP_QUERY ioctl(). This revision dropped the tool from the patch set and, once the API lands upstream, this tool might be added separately on Github as an example.
Based on discussion on earlier revisions of this patch set, it turned out that this ioctl() API is competitive with highly-optimized text-based pre-processing pattern that perf tool is using. Based on perf discussion, this revision adds more flexibility in specifying a subset of VMAs that are of interest. Now it's possible to specify desired permissions of VMAs (e.g., request only executable ones) and/or restrict to only a subset of VMAs that have file backing. This further improves the efficiency when using this new API thanks to more selective (executable VMAs only) querying.
In addition to a custom benchmarking tool, and experimental perf integration (available at [0]), Daniel Mueller has since also implemented an experimental integration into blazesym (see [1]), a library used for stack trace symbolization by our server fleet-wide profiler and another on-device profiler agent that runs on weaker ARM devices. The latter ARM-based device profiler is especially sensitive to performance, and so we benchmarked and compared text-based /proc/<pid>/maps solution to the equivalent one using PROCMAP_QUERY ioctl().
Results are very encouraging, giving us 5x improvement for end-to-end so-called "address normalization" pass, which is the part of the symbolization process that happens locally on ARM device, before being sent out for further heavier-weight processing on more powerful remote server. Note that this is not an artificial microbenchmark. It's a full end-to-end API call being measured with real-world data on real-world device.
TEXT-BASED
==========
Benchmarking main/normalize_process_no_build_ids_uncached_maps main/normalize_process_no_build_ids_uncached_maps
time: [49.777 µs 49.982 µs 50.250 µs]
IOCTL-BASED
===========
Benchmarking main/normalize_process_no_build_ids_uncached_maps main/normalize_process_no_build_ids_uncached_maps
time: [10.328 µs 10.391 µs 10.457 µs] change: [−79.453% −79.304% −79.166%] (p = 0.00 < 0.02) Performance has improved.
You can see above that we see the drop from 50µs down to 10µs for exactly the same amount of work, with the same data and target process.
With the aforementioned custom tool, we see about ~40x improvement (it might vary a bit, depending on a specific captured set of addresses). And even for perf-based benchmark it's on par or slightly ahead when using permission-based filtering (fetching only executable VMAs).
Earlier revisions attempted to use per-VMA locking, if kernel was compiled with CONFIG_PER_VMA_LOCK=y, but it turned out that anon_vma_name() is not yet compatible with per-VMA locking and assumes mmap_lock to be taken, which makes the use of per-VMA locking for this API premature. It was agreed ([2]) to continue for now with just mmap_lock, but the code structure is such that it should be easy to add per-VMA locking support once all the pieces are ready.
One thing that did not change was basing this new API as an ioctl() command on /proc/<pid>/maps file. An ioctl-based API on top of pidfd was considered, but has its own downsides. Implementing ioctl() directly on pidfd will cause access permission checks on every single ioctl(), which leads to performance concerns and potential spam of capable() audit messages. It also prevents a nice pattern, possible with /proc/<pid>/maps, in which application opens /proc/self/maps FD (requiring no additional capabilities) and passed this FD to profiling agent for querying. To achieve similar pattern, a new file would have to be created from pidf just for VMA querying, which is considered to be inferior to just querying /proc/<pid>/maps FD as proposed in current approach. These aspects were discussed in the hallway track at recent LSF/MM/BPF 2024 and sticking to procfs ioctl() was the final agreement we arrived at.
This patch set is based on top of mm-unstable branch in mm tree.
[0] https://github.com/anakryiko/linux/commits/procfs-proc-maps-ioctl-v2/
[1] https://github.com/libbpf/blazesym/pull/675
[2] https://lore.kernel.org/bpf/7rm3izyq2vjp5evdjc7c6z4crdd3oerpiknumdnmmemwyiwx7t@hleldw7iozi3/
v4->v5:
- added tests in selftests/proc (Andrew);
- added vma_page_size field (Liam);
- dropped the benchmark tool and BPF selftests parts, I'll send them
directly to bpf-next once this API makes it into that tree;
v3->v4:
- drop per-VMA locking changes for now, we'll need anon_vma_name() to be
compatible with vma->vm_lock approach (Suren, Liam);v2->v3:
- drop mmap_lock aggressively under CONFIG_PER_VMA_LOCK (Liam);
- code massaging to abstract per-VMA vs mmap_lock differences (Liam);v1->v2:
- per-VMA lock is used, if possible (Liam, Suren);
- added file-backed VMA querying (perf folks);
- added permission-based VMA querying (perf folks);
- split out build ID into separate patch (Suren);
- better documented API, added mention of ioctl() into procfs docs (Greg).
Andrii Nakryiko (6):
fs/procfs: extract logic for getting VMA name constituents fs/procfs: implement efficient VMA querying API for /proc/<pid>/maps fs/procfs: add build ID fetching to PROCMAP_QUERY API docs/procfs: call out ioctl()-based PROCMAP_QUERY command existence
tools: sync uapi/linux/fs.h header into tools subdir selftests/proc: add PROCMAP_QUERY ioctl tests
Documentation/filesystems/proc.rst | 9 + fs/proc/task_mmu.c | 383 ++++++++++++++++++--- include/uapi/linux/fs.h | 158 ++++++++- tools/include/uapi/linux/fs.h | 184 +++++++++- tools/testing/selftests/proc/Makefile | 1 + tools/testing/selftests/proc/proc-pid-vm.c | 86 +++++ 6 files changed, 754 insertions(+), 67 deletions(-)
Signature · 1 line
2.43.0
Andrii Nakryiko
maintainerPATCH v5 1/6 fs/procfs: extract logic for getting VMA name constituents
Extract generic logic to fetch relevant pieces of data to describe VMA name. This could be just some string (either special constant or user-provided), or a string with some formatted wrapping text (e.g., "[anon_shmem:<something>]"), or, commonly, file path. seq_file-based logic has different methods to handle all three cases, but they are currently mixed in with extracting underlying sources of data.
This patch splits this into data fetching and data formatting, so that data fetching can be reused later on.
There should be no functional changes.
--- fs/proc/task_mmu.c | 125 +++++++++++++++++++++++++-------------------- 1 file changed, 71 insertions(+), 54 deletions(-)
fs/proc/task_mmu.c +71 -54
@@ -239,6 +239,67 @@ static int do_maps_open(struct inode *inode, struct file *file, sizeof(struct proc_maps_private)); } +static void get_vma_name (struct vm_area_struct *vma, +const struct path **path, +const char **name, +const char **name_fmt) +{ +struct anon_vma_name *anon_name = vma->vm_mm ?anon_vma_name (vma) :NULL ; + + *name =NULL ; + *path =NULL ; + *name_fmt =NULL ; + +/* +* Print the dentry name for named mappings, and a +* special [heap] marker for the heap: +*/ +if (vma->vm_file) { +/* +* If user named this anon shared memory via +* prctl(PR_SET_VMA ..., use the provided name. +*/ +if (anon_name) { + *name_fmt ="[anon_shmem:%s]" ; + *name = anon_name->name; + }else { + *path =file_user_path (vma->vm_file); + } +return ; + } + +if (vma->vm_ops && vma->vm_ops->name) { + *name = vma->vm_ops->name (vma); +if (*name) +return ; + } + + *name =arch_vma_name (vma); +if (*name) +return ; + +if (!vma->vm_mm) { + *name ="[vdso]" ; +return ; + } + +if (vma_is_initial_heap (vma)) { + *name ="[heap]" ; +return ; + } + +if (vma_is_initial_stack (vma)) { + *name ="[stack]" ; +return ; + } + +if (anon_name) { + *name_fmt ="[anon:%s]" ; + *name = anon_name->name; +return ; + } +} + static void show_vma_header_prefix(struct seq_file *m, unsigned long start, unsigned long end, vm_flags_t flags, unsigned long long pgoff,
@@ -262,17 +323,15 @@ static void show_vma_header_prefix(struct seq_file *m,static void show_map_vma (struct seq_file *m,struct vm_area_struct *vma) { -struct anon_vma_name *anon_name =NULL ; -struct mm_struct *mm = vma->vm_mm; -struct file *file = vma->vm_file; +const struct path *path; +const char *name_fmt, *name;vm_flags_t flags = vma->vm_flags;unsigned long ino =0 ;unsigned long long pgoff =0 ;unsigned long start, end;dev_t dev =0 ; -const char *name =NULL ; -if (file) { +if (vma->vm_file) {const struct inode *inode =file_user_inode (vma->vm_file); dev = inode->i_sb->s_dev;
@@ -283,57 +342,15 @@ show_map_vma(struct seq_file *m, struct vm_area_struct *vma) start = vma->vm_start; end = vma->vm_end;show_vma_header_prefix (m, start, end, flags, pgoff, dev, ino); -if (mm) - anon_name =anon_vma_name (vma); -/* -* Print the dentry name for named mappings, and a -* special [heap] marker for the heap: -*/ -if (file) { +get_vma_name (vma, &path, &name, &name_fmt); +if (path) {seq_pad (m,' ' ); -/* -* If user named this anon shared memory via -* prctl(PR_SET_VMA ..., use the provided name. -*/ -if (anon_name) -seq_printf (m,"[anon_shmem:%s]" , anon_name->name); -else -seq_path (m,file_user_path (file),"\n" ); - goto done; - } - -if (vma->vm_ops && vma->vm_ops->name) { - name = vma->vm_ops->name (vma); -if (name) - goto done; - } - - name =arch_vma_name (vma); -if (!name) { -if (!mm) { - name ="[vdso]" ; - goto done; - } - -if (vma_is_initial_heap (vma)) { - name ="[heap]" ; - goto done; - } - -if (vma_is_initial_stack (vma)) { - name ="[stack]" ; - goto done; - } - -if (anon_name) { -seq_pad (m,' ' ); -seq_printf (m,"[anon:%s]" , anon_name->name); - } - } - -done: -if (name) { +seq_path (m, path,"\n" ); + }else if (name_fmt) { +seq_pad (m,' ' ); +seq_printf (m, name_fmt, name); + }else if (name) {seq_pad (m,' ' );seq_puts (m, name); }
Signature · 1 line
2.43.0
Andrii Nakryiko
maintainerPATCH v5 2/6 fs/procfs: implement efficient VMA querying API for /proc/<pid>/maps
In reply to Andrii Nakryiko/proc/<pid>/maps file is extremely useful in practice for various tasks involving figuring out process memory layout, what files are backing any given memory range, etc. One important class of applications that absolutely rely on this are profilers/stack symbolizers (perf tool being one of them). Patterns of use differ, but they generally would fall into two categories.
In on-demand pattern, a profiler/symbolizer would normally capture stack trace containing absolute memory addresses of some functions, and would then use /proc/<pid>/maps file to find corresponding backing ELF files (normally, only executable VMAs are of interest), file offsets within them, and then continue from there to get yet more information (ELF symbols, DWARF information) to get human-readable symbolic information. This pattern is used by Meta's fleet-wide profiler, as one example.
In preprocessing pattern, application doesn't know the set of addresses of interest, so it has to fetch all relevant VMAs (again, probably only executable ones), store or cache them, then proceed with profiling and stack trace capture. Once done, it would do symbolization based on stored VMA information. This can happen at much later point in time. This patterns is used by perf tool, as an example.
In either case, there are both performance and correctness requirement involved. This address to VMA information translation has to be done as efficiently as possible, but also not miss any VMA (especially in the case of loading/unloading shared libraries). In practice, correctness can't be guaranteed (due to process dying before VMA data can be captured, or shared library being unloaded, etc), but any effort to maximize the chance of finding the VMA is appreciated.
Unfortunately, for all the /proc/<pid>/maps file universality and usefulness, it doesn't fit the above use cases 100%.
First, it's main purpose is to emit all VMAs sequentially, but in practice captured addresses would fall only into a smaller subset of all process' VMAs, mainly containing executable text. Yet, library would need to parse most or all of the contents to find needed VMAs, as there is no way to skip VMAs that are of no use. Efficient library can do the linear pass and it is still relatively efficient, but it's definitely an overhead that can be avoided, if there was a way to do more targeted querying of the relevant VMA information.
Second, it's a text based interface, which makes its programmatic use from applications and libraries more cumbersome and inefficient due to the need to handle text parsing to get necessary pieces of information. The overhead is actually payed both by kernel, formatting originally binary VMA data into text, and then by user space application, parsing it back into binary data for further use.
For the on-demand pattern of usage, described above, another problem when writing generic stack trace symbolization library is an unfortunate performance-vs-correctness tradeoff that needs to be made. Library has to make a decision to either cache parsed contents of /proc/<pid>/maps (after initial processing) to service future requests (if application requests to symbolize another set of addresses (for the same process), captured at some later time, which is typical for periodic/continuous profiling cases) to avoid higher costs of re-parsing this file. Or it has to choose to cache the contents in memory to speed up future requests. In the former case, more memory is used for the cache and there is a risk of getting stale data if application loads or unloads shared libraries, or otherwise changed its set of VMAs somehow, e.g., through additional mmap() calls. In the latter case, it's the performance hit that comes from re-opening the file and re-parsing its contents all over again.
This patch aims to solve this problem by providing a new API built on top of /proc/<pid>/maps. It's meant to address both non-selectiveness and text nature of /proc/<pid>/maps, by giving user more control of what sort of VMA(s) needs to be queried, and being binary-based interface eliminates the overhead of text formatting (on kernel side) and parsing (on user space side).
It's also designed to be extensible and forward/backward compatible by including required struct size field, which user has to provide. We use established copy_struct_from_user() approach to handle extensibility.
User has a choice to pick either getting VMA that covers provided address or -ENOENT if none is found (exact, least surprising, case). Or, with an extra query flag (PROCMAP_QUERY_COVERING_OR_NEXT_VMA), they can get either VMA that covers the address (if there is one), or the closest next VMA (i.e., VMA with the smallest vm_start > addr). The latter allows more efficient use, but, given it could be a surprising behavior, requires an explicit opt-in.
There is another query flag that is useful for some use cases. PROCMAP_QUERY_FILE_BACKED_VMA instructs this API to only return file-backed VMAs. Combining this with PROCMAP_QUERY_COVERING_OR_NEXT_VMA makes it possible to efficiently iterate only file-backed VMAs of the process, which is what profilers/symbolizers are normally interested in.
All the above querying flags can be combined with (also optional) set of desired VMA permissions flags. This allows to, for example, iterate only an executable subset of VMAs, which is what preprocessing pattern, used by perf tool, would benefit from, as the assumption is that captured stack traces would have addresses of executable code. This saves time by skipping non-executable VMAs altogether efficienty.
All these querying flags (modifiers) are orthogonal and can be combined in a semantically meaningful and natural way.
Basing this ioctl()-based API on top of /proc/<pid>/maps's FD makes sense given it's querying the same set of VMA data. It's also benefitial because permission checks for /proc/<pid>/maps is performed at open time once, and the actual data read of text contents of /proc/<pid>/maps is done without further permission checks. We piggyback on this pattern with ioctl()-based API as well, as that's a desired property. Both for performance reasons, but also for security and flexibility reasons.
Allowing application to open an FD for /proc/self/maps without any extra capabilities, and then passing it to some sort of profiling agent through Unix-domain socket, would allow such profiling agent to not require some of the capabilities that are otherwise expected when opening /proc/<pid>/maps file for *another* process. This is a desirable property for some more restricted setups.
This new ioctl-based implementation doesn't interfere with seq_file-based implementation of /proc/<pid>/maps textual interface, and so could be used together or independently without paying any price for that.
Note also, that fetching VMA name (e.g., backing file path, or special hard-coded or user-provided names) is optional just like build ID. If user sets vma_name_size to zero, kernel code won't attempt to retrieve it, saving resources.
Earlier versions of this patch set were adding per-VMA locking, which is why we have a code structure that is ready for abstracting mmap_lock vs vm_lock differences (query_vma_setup(), query_vma_teardown(), and query_vma_find_by_addr()), but given anon_vma_name() is not yet compatible with per-VMA locking, initial implementation sticks to using only mmap_lock for now. It will be easy to add back per-VMA locking once all the pieces are ready later on. Which is why we keep existing code structure with setup/teardown/query helper functions.
--- fs/proc/task_mmu.c | 235 ++++++++++++++++++++++++++++++++++++++++ include/uapi/linux/fs.h | 130 +++++++++++++++++++++- 2 files changed, 364 insertions(+), 1 deletion(-)
fs/proc/task_mmu.c +235 -0
@@ -375,11 +375,246 @@ static int pid_maps_open(struct inode *inode, struct file *file)return do_maps_open (inode, file, &proc_pid_maps_op); } +#define PROCMAP_QUERY_VMA_FLAGS ( \ + PROCMAP_QUERY_VMA_READABLE | \ + PROCMAP_QUERY_VMA_WRITABLE | \ + PROCMAP_QUERY_VMA_EXECUTABLE | \ + PROCMAP_QUERY_VMA_SHARED \ +) + +#define PROCMAP_QUERY_VALID_FLAGS_MASK ( \ + PROCMAP_QUERY_COVERING_OR_NEXT_VMA | \ + PROCMAP_QUERY_FILE_BACKED_VMA | \ + PROCMAP_QUERY_VMA_FLAGS \ +) + +static int query_vma_setup (struct mm_struct *mm) +{ +return mmap_read_lock_killable (mm); +} + +static void query_vma_teardown (struct mm_struct *mm,struct vm_area_struct *vma) +{ +mmap_read_unlock (mm); +} + +static struct vm_area_struct *query_vma_find_by_addr (struct mm_struct *mm,unsigned long addr) +{ +return find_vma (mm, addr); +} + +static struct vm_area_struct *query_matching_vma (struct mm_struct *mm, +unsigned long addr,u32 flags) +{ +struct vm_area_struct *vma; + +next_vma: + vma =query_vma_find_by_addr (mm, addr); +if (!vma) + goto no_vma; + +/* user requested only file-backed VMA, keep iterating */ +if ((flags &PROCMAP_QUERY_FILE_BACKED_VMA ) && !vma->vm_file) + goto skip_vma; + +/* VMA permissions should satisfy query flags */ +if (flags &PROCMAP_QUERY_VMA_FLAGS ) { +u32 perm =0 ; + +if (flags &PROCMAP_QUERY_VMA_READABLE ) + perm |=VM_READ ; +if (flags &PROCMAP_QUERY_VMA_WRITABLE ) + perm |=VM_WRITE ; +if (flags &PROCMAP_QUERY_VMA_EXECUTABLE ) + perm |=VM_EXEC ; +if (flags &PROCMAP_QUERY_VMA_SHARED ) + perm |=VM_MAYSHARE ; + +if ((vma->vm_flags & perm) != perm) + goto skip_vma; + } + +/* found covering VMA or user is OK with the matching next VMA */ +if ((flags &PROCMAP_QUERY_COVERING_OR_NEXT_VMA ) || vma->vm_start <= addr) +return vma; + +skip_vma: +/* +* If the user needs closest matching VMA, keep iterating. +*/ + addr = vma->vm_end; +if (flags &PROCMAP_QUERY_COVERING_OR_NEXT_VMA ) + goto next_vma; +no_vma: +return ERR_PTR (-ENOENT ); +} + +static int do_procmap_query (struct proc_maps_private *priv,void __user *uarg) +{ +struct procmap_query karg; +struct vm_area_struct *vma; +struct mm_struct *mm; +const char *name =NULL ; +char *name_buf =NULL ; +__u64 usize; +int err; + +if (copy_from_user (&usize, (void __user *)uarg,sizeof (usize))) +return -EFAULT ; +/* argument struct can never be that large, reject abuse */ +if (usize >PAGE_SIZE ) +return -E2BIG ; +/* argument struct should have at least query_flags and query_addr fields */ + if (usize < offsetofend(struct procmap_query, query_addr)) +return -EINVAL ; + err =copy_struct_from_user (&karg,sizeof (karg), uarg, usize); +if (err) +return err; + +/* reject unknown flags */ +if (karg.query_flags & ~PROCMAP_QUERY_VALID_FLAGS_MASK ) +return -EINVAL ; +/* either both buffer address and size are set, or both should be zero */ +if (!!karg.vma_name_size != !!karg.vma_name_addr) +return -EINVAL ; + + mm = priv->mm; +if (!mm || !mmget_not_zero (mm)) +return -ESRCH ; + + err =query_vma_setup (mm); +if (err) { +mmput (mm); +return err; + } + + vma =query_matching_vma (mm, karg.query_addr, karg.query_flags); +if (IS_ERR (vma)) { + err =PTR_ERR (vma); + vma =NULL ; + goto out; + } + + karg.vma_start = vma->vm_start; + karg.vma_end = vma->vm_end; + + karg.vma_flags =0 ; +if (vma->vm_flags &VM_READ ) + karg.vma_flags |=PROCMAP_QUERY_VMA_READABLE ; +if (vma->vm_flags &VM_WRITE ) + karg.vma_flags |=PROCMAP_QUERY_VMA_WRITABLE ; +if (vma->vm_flags &VM_EXEC ) + karg.vma_flags |=PROCMAP_QUERY_VMA_EXECUTABLE ; +if (vma->vm_flags &VM_MAYSHARE ) + karg.vma_flags |=PROCMAP_QUERY_VMA_SHARED ; + + karg.vma_page_size =vma_kernel_pagesize (vma); + +if (vma->vm_file) { +const struct inode *inode =file_user_inode (vma->vm_file); + + karg.vma_offset = ((__u64 )vma->vm_pgoff) <<PAGE_SHIFT ; + karg.dev_major =MAJOR (inode->i_sb->s_dev); + karg.dev_minor =MINOR (inode->i_sb->s_dev); + karg.inode = inode->i_ino; + }else { + karg.vma_offset =0 ; + karg.dev_major =0 ; + karg.dev_minor =0 ; + karg.inode =0 ; + } + +if (karg.build_id_size) { +__u32 build_id_sz; + + err =build_id_parse (vma, build_id_buf, &build_id_sz); +if (err) { + karg.build_id_size =0 ; + }else { +if (karg.build_id_size < build_id_sz) { + err = -ENAMETOOLONG ; + goto out; + } + karg.build_id_size = build_id_sz; + } + } + +if (karg.vma_name_size) { +size_t name_buf_sz =min_t (size_t,PATH_MAX , karg.vma_name_size); +const struct path *path; +const char *name_fmt; +size_t name_sz =0 ; + +get_vma_name (vma, &path, &name, &name_fmt); + +if (path || name_fmt || name) { + name_buf =kmalloc (name_buf_sz,GFP_KERNEL ); +if (!name_buf) { + err = -ENOMEM ; + goto out; + } + } +if (path) { + name =d_path (path, name_buf, name_buf_sz); +if (IS_ERR (name)) { + err =PTR_ERR (name); + goto out; + } + name_sz = name_buf + name_buf_sz - name; + }else if (name || name_fmt) { + name_sz =1 +snprintf (name_buf, name_buf_sz, name_fmt ?:"%s" , name); + name = name_buf; + } +if (name_sz > name_buf_sz) { + err = -ENAMETOOLONG ; + goto out; + } + karg.vma_name_size = name_sz; + } + +/* unlock vma or mmap_lock, and put mm_struct before copying data to user */ +query_vma_teardown (mm, vma); +mmput (mm); + +if (karg.vma_name_size &©_to_user ((void __user *)karg.vma_name_addr, + name, karg.vma_name_size)) { +kfree (name_buf); +return -EFAULT ; + } +kfree (name_buf); + +if (copy_to_user (uarg, &karg,min_t (size_t,sizeof (karg), usize))) +return -EFAULT ; + +return 0 ; + +out: +query_vma_teardown (mm, vma); +mmput (mm); +kfree (name_buf); +return err; +} + +static long procfs_procmap_ioctl (struct file *file,unsigned int cmd,unsigned long arg) +{ +struct seq_file *seq = file->private_data; +struct proc_maps_private *priv = seq->private; + +switch (cmd) { +case PROCMAP_QUERY : +return do_procmap_query (priv, (void __user *)arg); +default : +return -ENOIOCTLCMD ; + } +} +const struct file_operations proc_pid_maps_operations = { .open = pid_maps_open, .read = seq_read, .llseek = seq_lseek, .release = proc_map_release, + .unlocked_ioctl = procfs_procmap_ioctl, + .compat_ioctl = procfs_procmap_ioctl, };/*
include/uapi/linux/fs.h +129 -1
141 lines not on this page · raw message
Signature · 1 line
2.43.0
Re: [PATCH v5 2/6] fs/procfs: implement efficient VMA querying API for /proc/<pid>/maps
* Andrii Nakryiko <[email protected]> [240618 18:45]:
...
Quoted 62 lines
fs/proc/task_mmu.c +62
-0+ +static int do_procmap_query (struct proc_maps_private *priv,void __user *uarg) +{ +struct procmap_query karg; +struct vm_area_struct *vma; +struct mm_struct *mm; +const char *name =NULL ; +char *name_buf =NULL ; +__u64 usize; +int err; + +if (copy_from_user (&usize, (void __user *)uarg,sizeof (usize))) +return -EFAULT ; +/* argument struct can never be that large, reject abuse */ +if (usize >PAGE_SIZE ) +return -E2BIG ; +/* argument struct should have at least query_flags and query_addr fields */ + if (usize < offsetofend(struct procmap_query, query_addr)) +return -EINVAL ; + err =copy_struct_from_user (&karg,sizeof (karg), uarg, usize); +if (err) +return err; + +/* reject unknown flags */ +if (karg.query_flags & ~PROCMAP_QUERY_VALID_FLAGS_MASK ) +return -EINVAL ; +/* either both buffer address and size are set, or both should be zero */ +if (!!karg.vma_name_size != !!karg.vma_name_addr) +return -EINVAL ; + + mm = priv->mm; +if (!mm || !mmget_not_zero (mm)) +return -ESRCH ; + + err =query_vma_setup (mm); +if (err) { +mmput (mm); +return err; + } + + vma =query_matching_vma (mm, karg.query_addr, karg.query_flags); +if (IS_ERR (vma)) { + err =PTR_ERR (vma); + vma =NULL ; + goto out; + } + + karg.vma_start = vma->vm_start; + karg.vma_end = vma->vm_end; + + karg.vma_flags =0 ; +if (vma->vm_flags &VM_READ ) + karg.vma_flags |=PROCMAP_QUERY_VMA_READABLE ; +if (vma->vm_flags &VM_WRITE ) + karg.vma_flags |=PROCMAP_QUERY_VMA_WRITABLE ; +if (vma->vm_flags &VM_EXEC ) + karg.vma_flags |=PROCMAP_QUERY_VMA_EXECUTABLE ; +if (vma->vm_flags &VM_MAYSHARE ) + karg.vma_flags |=PROCMAP_QUERY_VMA_SHARED ; + + karg.vma_page_size =vma_kernel_pagesize (vma); +
...
Quoted 38 lines
include/uapi/linux/fs.h +38
-0+/* +* Input/output argument structured passed into ioctl() call. It can be used +* to query a set of VMAs (Virtual Memory Areas) of a process. +* +* Each field can be one of three kinds, marked in a short comment to the +* right of the field: +* - "in", input argument, user has to provide this value, kernel doesn't modify it; +* - "out", output argument, kernel sets this field with VMA data; +* - "in/out", input and output argument; user provides initial value (used +* to specify maximum allowable buffer size), and kernel sets it to actual +* amount of data written (or zero, if there is no data). +* +* If matching VMA is found (according to criterias specified by +* query_addr/query_flags, all the out fields are filled out, and ioctl() +* returns 0. If there is no matching VMA, -ENOENT will be returned. +* In case of any other error, negative error code other than -ENOENT is +* returned. +* +* Most of the data is similar to the one returned as text in /proc/<pid>/maps +* file, but procmap_query provides more querying flexibility. There are no +* consistency guarantees between subsequent ioctl() calls, but data returned +* for matched VMA is self-consistent. +*/ +struct procmap_query { +/* Query struct size, for backwards/forward compatibility */ +__u64 size; +/* +* Query flags, a combination of enum procmap_query_flags values. +* Defines query filtering and behavior, see enum procmap_query_flags. +* +* Input argument, provided by user. Kernel doesn't modify it. +*/ +__u64 query_flags;/* in */ +/* +* Query address. By default, VMA that covers this address will +* be looked up. PROCMAP_QUERY_* flags above modify this default +* behavior further. +*
include/uapi/linux/fs.h struct pm_scan_arg { · line 482
Quoted 10 lines
include/uapi/linux/fs.h +10
-0+* Input argument, provided by user. Kernel doesn't modify it. +*/ +__u64 query_addr;/* in */ +/* VMA starting (inclusive) and ending (exclusive) address, if VMA is found. */ +__u64 vma_start;/* out */ +__u64 vma_end;/* out */ +/* VMA permissions flags. A combination of PROCMAP_QUERY_VMA_* flags. */ +__u64 vma_flags;/* out */ +/* VMA backing page size granularity. */ +__u32 vma_page_size;/* out */
The vma_kernel_pagesize() returns an unsigned long. We could potentially be truncating the returned value (although probably not today?). This is from the vm_operations_struct pagesize, which also returns an unsigned long. Could we switch this to __u64?
...
Thanks,
Liam
Re: [PATCH v5 2/6] fs/procfs: implement efficient VMA querying API for /proc/<pid>/maps
Quoted 113 lines · Liam R. Howlett
* Andrii Nakryiko <[email protected]> [240618 18:45]:
...Quoted 62 lines
fs/proc/task_mmu.c +62
-0+ +static int do_procmap_query (struct proc_maps_private *priv,void __user *uarg) +{ +struct procmap_query karg; +struct vm_area_struct *vma; +struct mm_struct *mm; +const char *name =NULL ; +char *name_buf =NULL ; +__u64 usize; +int err; + +if (copy_from_user (&usize, (void __user *)uarg,sizeof (usize))) +return -EFAULT ; +/* argument struct can never be that large, reject abuse */ +if (usize >PAGE_SIZE ) +return -E2BIG ; +/* argument struct should have at least query_flags and query_addr fields */ + if (usize < offsetofend(struct procmap_query, query_addr)) +return -EINVAL ; + err =copy_struct_from_user (&karg,sizeof (karg), uarg, usize); +if (err) +return err; + +/* reject unknown flags */ +if (karg.query_flags & ~PROCMAP_QUERY_VALID_FLAGS_MASK ) +return -EINVAL ; +/* either both buffer address and size are set, or both should be zero */ +if (!!karg.vma_name_size != !!karg.vma_name_addr) +return -EINVAL ; + + mm = priv->mm; +if (!mm || !mmget_not_zero (mm)) +return -ESRCH ; + + err =query_vma_setup (mm); +if (err) { +mmput (mm); +return err; + } + + vma =query_matching_vma (mm, karg.query_addr, karg.query_flags); +if (IS_ERR (vma)) { + err =PTR_ERR (vma); + vma =NULL ; + goto out; + } + + karg.vma_start = vma->vm_start; + karg.vma_end = vma->vm_end; + + karg.vma_flags =0 ; +if (vma->vm_flags &VM_READ ) + karg.vma_flags |=PROCMAP_QUERY_VMA_READABLE ; +if (vma->vm_flags &VM_WRITE ) + karg.vma_flags |=PROCMAP_QUERY_VMA_WRITABLE ; +if (vma->vm_flags &VM_EXEC ) + karg.vma_flags |=PROCMAP_QUERY_VMA_EXECUTABLE ; +if (vma->vm_flags &VM_MAYSHARE ) + karg.vma_flags |=PROCMAP_QUERY_VMA_SHARED ; + + karg.vma_page_size =vma_kernel_pagesize (vma); +...
Quoted 45 lines
include/uapi/linux/fs.h +45
-0+/* +* Input/output argument structured passed into ioctl() call. It can be used +* to query a set of VMAs (Virtual Memory Areas) of a process. +* +* Each field can be one of three kinds, marked in a short comment to the +* right of the field: +* - "in", input argument, user has to provide this value, kernel doesn't modify it; +* - "out", output argument, kernel sets this field with VMA data; +* - "in/out", input and output argument; user provides initial value (used +* to specify maximum allowable buffer size), and kernel sets it to actual +* amount of data written (or zero, if there is no data). +* +* If matching VMA is found (according to criterias specified by +* query_addr/query_flags, all the out fields are filled out, and ioctl() +* returns 0. If there is no matching VMA, -ENOENT will be returned. +* In case of any other error, negative error code other than -ENOENT is +* returned. +* +* Most of the data is similar to the one returned as text in /proc/<pid>/maps +* file, but procmap_query provides more querying flexibility. There are no +* consistency guarantees between subsequent ioctl() calls, but data returned +* for matched VMA is self-consistent. +*/ +struct procmap_query { +/* Query struct size, for backwards/forward compatibility */ +__u64 size; +/* +* Query flags, a combination of enum procmap_query_flags values. +* Defines query filtering and behavior, see enum procmap_query_flags. +* +* Input argument, provided by user. Kernel doesn't modify it. +*/ +__u64 query_flags;/* in */ +/* +* Query address. By default, VMA that covers this address will +* be looked up. PROCMAP_QUERY_* flags above modify this default +* behavior further. +* +* Input argument, provided by user. Kernel doesn't modify it. +*/ +__u64 query_addr;/* in */ +/* VMA starting (inclusive) and ending (exclusive) address, if VMA is found. */ +__u64 vma_start;/* out */ +__u64 vma_end;/* out */ +/* VMA permissions flags. A combination of PROCMAP_QUERY_VMA_* flags. */
Quoted 9 lines
include/uapi/linux/fs.h struct pm_scan_arg { · line 482
Quoted 3 lines
include/uapi/linux/fs.h +3
-0+__u64 vma_flags;/* out */ +/* VMA backing page size granularity. */ +__u32 vma_page_size;/* out */ The
vma_kernel_pagesize()returns an unsigned long. We could potentially be truncating the returned value (although probably not today?). This is from thevm_operations_structpagesize, which also returns an unsigned long. Could we switch this to__u64?
Yep, fair point. It seemed excessive to use u64, but I guess nothing prevents us (in the future) from having humongous pages with >4GB sizes. I'll update to __u64.
I'll wait for a few more hours just in case I get some other feedback, and will rebase and post a new revision later today, thanks.
Quoted 4 lines
...
Thanks,
Liam
Andrii Nakryiko
maintainerPATCH v5 3/6 fs/procfs: add build ID fetching to PROCMAP_QUERY API
In reply to Andrii NakryikoThe need to get ELF build ID reliably is an important aspect when dealing with profiling and stack trace symbolization, and /proc/<pid>/maps textual representation doesn't help with this.
To get backing file's ELF build ID, application has to first resolve VMA, then use it's start/end address range to follow a special /proc/<pid>/map_files/<start>-<end> symlink to open the ELF file (this is necessary because backing file might have been removed from the disk or was already replaced with another binary in the same file path.
Such approach, beyond just adding complexity of having to do a bunch of extra work, has extra security implications. Because application opens underlying ELF file and needs read access to its entire contents (as far as kernel is concerned), kernel puts additional capable() checks on following /proc/<pid>/map_files/<start>-<end> symlink. And that makes sense in general.
But in the case of build ID, profiler/symbolizer doesn't need the contents of ELF file, per se. It's only build ID that is of interest, and ELF build ID itself doesn't provide any sensitive information.
So this patch adds a way to request backing file's ELF build ID along the rest of VMA information in the same API. User has control over whether this piece of information is requested or not by either setting build_id_size field to zero or non-zero maximum buffer size they provided through build_id_addr field (which encodes user pointer as __u64 field). This is a completely optional piece of information, and so has no performance implications for user cases that don't care about build ID, while improving performance and simplifying the setup for those application that do need it.
Kernel already implements build ID fetching, which is used from BPF subsystem. We are reusing this code here, but plan a follow up changes to make it work better under more relaxed assumption (compared to what existing code assumes) of being called from user process context, in which page faults are allowed. BPF-specific implementation currently bails out if necessary part of ELF file is not paged in, all due to extra BPF-specific restrictions (like the need to fetch build ID in restrictive contexts such as NMI handler).
--- fs/proc/task_mmu.c | 25 ++++++++++++++++++++++++- include/uapi/linux/fs.h | 28 ++++++++++++++++++++++++++++ 2 files changed, 52 insertions(+), 1 deletion(-)
fs/proc/task_mmu.c +24 -1
@@ -22,6 +22,7 @@#include <linux/pkeys.h> #include <linux/minmax.h> #include <linux/overflow.h> +#include <linux/buildid.h> #include <asm/elf.h> #include <asm/tlb.h>
@@ -445,6 +446,7 @@ static struct vm_area_struct *query_matching_vma(struct mm_struct *mm, addr = vma->vm_end;if (flags &PROCMAP_QUERY_COVERING_OR_NEXT_VMA ) goto next_vma; + no_vma:return ERR_PTR (-ENOENT ); }
@@ -455,7 +457,7 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg)struct vm_area_struct *vma;struct mm_struct *mm;const char *name =NULL ; -char *name_buf =NULL ; +char build_id_buf[BUILD_ID_SIZE_MAX ], *name_buf =NULL ;__u64 usize;int err;
@@ -477,6 +479,8 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg)/* either both buffer address and size are set, or both should be zero */ if (!!karg.vma_name_size != !!karg.vma_name_addr)return -EINVAL ; +if (!!karg.build_id_size != !!karg.build_id_addr) +return -EINVAL ; mm = priv->mm;if (!mm || !mmget_not_zero (mm))
@@ -539,6 +543,21 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg) } } +if (karg.build_id_size) { +__u32 build_id_sz; + + err =build_id_parse (vma, build_id_buf, &build_id_sz); +if (err) { + karg.build_id_size =0 ; + }else { +if (karg.build_id_size < build_id_sz) { + err = -ENAMETOOLONG ; + goto out; + } + karg.build_id_size = build_id_sz; + } + } +if (karg.vma_name_size) {size_t name_buf_sz =min_t (size_t,PATH_MAX , karg.vma_name_size);const struct path *path;
@@ -583,6 +602,10 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg) }kfree (name_buf); +if (karg.build_id_size &©_to_user ((void __user *)karg.build_id_addr, + build_id_buf, karg.build_id_size)) +return -EFAULT ; +if (copy_to_user (uarg, &karg,min_t (size_t,sizeof (karg), usize)))return -EFAULT ;
include/uapi/linux/fs.h +28 -0
@@ -511,6 +511,26 @@ struct procmap_query {* If set to zero, vma_name_addr should be set to zero as well */ __u32 vma_name_size;/* in/out */ +/* +* If set to non-zero value, signals the request to extract and return +* VMA's backing file's build ID, if the backing file is an ELF file +* and it contains embedded build ID. +* +* Kernel will set this field to zero, if VMA has no backing file, +* backing file is not an ELF file, or ELF file has no build ID +* embedded. +* +* Build ID is a binary value (not a string). Kernel will set +* build_id_size field to exact number of bytes used for build ID. +* If build ID is requested and present, but needs more bytes than +* user-supplied maximum buffer size (see build_id_addr field below), +* -E2BIG error will be returned. +* +* If this field is set to non-zero value, build_id_addr should point +* to valid user space memory buffer of at least build_id_size bytes. +* If set to zero, build_id_addr should be set to zero as well +*/ +__u32 build_id_size;/* in/out */ /* * User-supplied address of a buffer of at least vma_name_size bytes * for kernel to fill with matched VMA's name (see vma_name_size field
@@ -519,6 +539,14 @@ struct procmap_query {* Should be set to zero if VMA name should not be returned. */ __u64 vma_name_addr;/* in */ +/* +* User-supplied address of a buffer of at least build_id_size bytes +* for kernel to fill with matched VMA's ELF build ID, if available +* (see build_id_size field description above for details). +* +* Should be set to zero if build ID should not be returned. +*/ + __u64 build_id_addr; /* in */ };#endif /* _UAPI_LINUX_FS_H */
Signature · 1 line
2.43.0
Re: [PATCH v5 3/6] fs/procfs: add build ID fetching to PROCMAP_QUERY API
fs/proc/task_mmu.c static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg) · line 549
Quoted 12 lines · Andrii Nakryiko
The need to get ELF build ID reliably is an important aspect when dealing with profiling and stack trace symbolization, and /proc/<pid>/maps textual representation doesn't help with this.
fs/proc/task_mmu.c +4
-0@@ -539,6 +543,21 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg) } } +if (karg.build_id_size) { +__u32 build_id_sz; + + err =build_id_parse (vma, build_id_buf, &build_id_sz);
This is not your bug but build_id_parse() assumes program headers immediately follow ELF header which is not guaranteed.
Quoted 10 lines
include/uapi/linux/fs.h +5
-0+* If this field is set to non-zero value, build_id_addr should point +* to valid user space memory buffer of at least build_id_size bytes. +* If set to zero, build_id_addr should be set to zero as well +*/ +__u32 build_id_size;/* in/out */ /* * User-supplied address of a buffer of at least vma_name_size bytes * for kernel to fill with matched VMA's name (see vma_name_size field @@ -519,6 +539,14 @@ struct procmap_query {* Should be set to zero if VMA name should not be returned.
include/uapi/linux/fs.h struct procmap_query { · line 549
Quoted 10 lines
include/uapi/linux/fs.h +8
-0*/ __u64 vma_name_addr;/* in */ +/* +* User-supplied address of a buffer of at least build_id_size bytes +* for kernel to fill with matched VMA's ELF build ID, if available +* (see build_id_size field description above for details). +* +* Should be set to zero if build ID should not be returned. +*/ +__u64 build_id_addr;/* in */
Can this be simplified to 512-bit buffer in ioctl structure? BUILD_ID_SIZE_MAX is 20 which is sha1.
Re: [PATCH v5 3/6] fs/procfs: add build ID fetching to PROCMAP_QUERY API
Quoted 10 lines · Alexey Dobriyan
Quoted 8 lines · Andrii Nakryiko
The need to get ELF build ID reliably is an important aspect when dealing with profiling and stack trace symbolization, and /proc/<pid>/maps textual representation doesn't help with this.
(quoted diff) +0
-0@@ -539,6 +543,21 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg) } }
Quoted 7 lines
fs/proc/task_mmu.c static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg) · line 549
Quoted 4 lines
fs/proc/task_mmu.c +4
-0+if (karg.build_id_size) { +__u32 build_id_sz; + + err =build_id_parse (vma, build_id_buf, &build_id_sz);This is not your bug but
build_id_parse()assumes program headers immediately follow ELF header which is not guaranteed.
Yes, I'm aware, and I think I stated somewhere that I want to fix/improve that. The thing is, current build_id_parse() was built for BPF under NMI context assumption, which is why it can't page in memory and so on (and this "build ID has to be in the first page" was a surprise to me, but probably just a technical shortcut to make it a bit easier to implement). Regardless, my plan, once this API is merged, is to follow up with make build_id_parse() variant that would work reliably under sleepable context assumptions. Hopefully that's ok not to bundle all that with these patches?
Quoted 16 lines
Quoted 16 lines
include/uapi/linux/fs.h +8
-0+* If this field is set to non-zero value, build_id_addr should point +* to valid user space memory buffer of at least build_id_size bytes. +* If set to zero, build_id_addr should be set to zero as well +*/ +__u32 build_id_size;/* in/out */ /* * User-supplied address of a buffer of at least vma_name_size bytes * for kernel to fill with matched VMA's name (see vma_name_size field @@ -519,6 +539,14 @@ struct procmap_query {* Should be set to zero if VMA name should not be returned. */ __u64 vma_name_addr;/* in */ +/* +* User-supplied address of a buffer of at least build_id_size bytes +* for kernel to fill with matched VMA's ELF build ID, if available
Quoted 8 lines
include/uapi/linux/fs.h struct procmap_query { · line 549
Quoted 5 lines
include/uapi/linux/fs.h +5
-0+* (see build_id_size field description above for details). +* +* Should be set to zero if build ID should not be returned. +*/ +__u64 build_id_addr;/* in */ Can this be simplified to 512-bit buffer in ioctl structure?
BUILD_ID_SIZE_MAXis 20 which is sha1.
I'd prefer not to because vma_name can't use the same trick, so we'll have to have this size+buffer address approach anyway. And because of that I'd like to have all these optional variable-length/string output arguments handled in a uniform way. In practice, it's really simple to use this from user-space, the only mildly annoying part is casting pointer to __u64. But as I said, for vma_name users will do this anyways, so not much benefit simplifying the build_id part only.
Andrii Nakryiko
maintainerPATCH v5 4/6 docs/procfs: call out ioctl()-based PROCMAP_QUERY command existence
In reply to Andrii NakryikoCall out PROCMAP_QUERY ioctl() existence in the section describing /proc/PID/maps file in documentation. We refer user to UAPI header for low-level details of this programmatic interface.
--- Documentation/filesystems/proc.rst | 9 +++++++++ 1 file changed, 9 insertions(+)
Documentation/filesystems/proc.rst +9 -0
@@ -443,6 +443,15 @@ is not associated with a file: or if empty, the mapping is anonymous. +Starting with 6.11 kernel, /proc/PID/maps provides an alternative +ioctl()-based API that gives ability to flexibly and efficiently query and +filter individual VMAs. This interface is binary and is meant for more +efficient and easy programmatic use. `struct procmap_query`, defined in +linux/fs.h UAPI header, serves as an input/output argument to the +`PROCMAP_QUERY` ioctl() command. See comments in linus/fs.h UAPI header for +details on query semantics, supported flags, data returned, and general API +usage information. + The /proc/PID/smaps is an extension based on maps, showing the memory consumption for each of the process's mappings. For each mapping (aka Virtual Memory Area, or VMA) there is a series of lines such as the following::
Signature · 1 line
2.43.0
Andrii Nakryiko
maintainerPATCH v5 5/6 tools: sync uapi/linux/fs.h header into tools subdir
In reply to Andrii NakryikoWe need this UAPI header in tools/include subdirectory for using it from BPF selftests.
--- tools/include/uapi/linux/fs.h | 184 +++++++++++++++++++++++++++++++--- 1 file changed, 172 insertions(+), 12 deletions(-)
tools/include/uapi/linux/fs.h +172 -12
@@ -1,6 +1,6 @@/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ -#ifndef _LINUX_FS_H -#define _LINUX_FS_H +#ifndef _UAPI_LINUX_FS_H +#define _UAPI_LINUX_FS_H/* * This file has definitions for some important file table structures
@@ -13,10 +13,14 @@#include <linux/limits.h> #include <linux/ioctl.h> #include <linux/types.h> +#ifndef __KERNEL__#include <linux/fscrypt.h> +#endif /* Use of MS_* flags within the kernel is restricted to core mount(2) code. */ +#if !defined(__KERNEL__)#include <linux/mount.h> +#endif /* * It's silly to have NR_OPEN bigger than NR_FILE, but you can change
@@ -24,8 +28,8 @@* nr_file rlimit, so it's safe to set up a ridiculously high absolute * upper limit on files-per-process. * -* Some programs (notably those using select()) may have to be -* recompiled to take full advantage of the new limits.. +* Some programs (notably those using select()) may have to be +* recompiled to take full advantage of the new limits.. */ /* Fixed constants first: */
@@ -308,29 +312,31 @@ struct fsxattr {typedef int __bitwise __kernel_rwf_t ;/* high priority request, poll if possible */ -#define RWF_HIPRI ((__kernel_rwf_t)0x00000001) +#define RWF_HIPRI ((__force __kernel_rwf_t)0x00000001)/* per-IO O_DSYNC */ -#define RWF_DSYNC ((__kernel_rwf_t)0x00000002) +#define RWF_DSYNC ((__force __kernel_rwf_t)0x00000002)/* per-IO O_SYNC */ -#define RWF_SYNC ((__kernel_rwf_t)0x00000004) +#define RWF_SYNC ((__force __kernel_rwf_t)0x00000004)/* per-IO, return -EAGAIN if operation would block */ -#define RWF_NOWAIT ((__kernel_rwf_t)0x00000008) +#define RWF_NOWAIT ((__force __kernel_rwf_t)0x00000008)/* per-IO O_APPEND */ -#define RWF_APPEND ((__kernel_rwf_t)0x00000010) +#define RWF_APPEND ((__force __kernel_rwf_t)0x00000010)/* per-IO negation of O_APPEND */ -#define RWF_NOAPPEND ((__kernel_rwf_t)0x00000020) +#define RWF_NOAPPEND ((__force __kernel_rwf_t)0x00000020)/* mask of flags supported by the kernel */ #define RWF_SUPPORTED (RWF_HIPRI | RWF_DSYNC | RWF_SYNC | RWF_NOWAIT |\ RWF_APPEND | RWF_NOAPPEND) +#define PROCFS_IOCTL_MAGIC 'f' +/* Pagemap ioctl */ -#define PAGEMAP_SCAN _IOWR('f', 16, struct pm_scan_arg) +#define PAGEMAP_SCAN _IOWR(PROCFS_IOCTL_MAGIC, 16, struct pm_scan_arg)/* Bitmasks provided in pm_scan_args masks and reported in page_region.categories. */ #define PAGE_IS_WPALLOWED (1 << 0)
@@ -389,4 +395,158 @@ struct pm_scan_arg { __u64 return_mask; }; -#endif /* _LINUX_FS_H */ +/* /proc/<pid>/maps ioctl */ +#define PROCMAP_QUERY _IOWR(PROCFS_IOCTL_MAGIC, 17, struct procmap_query) + +enum procmap_query_flags { +/* +* VMA permission flags. +* +* Can be used as part of procmap_query.query_flags field to look up +* only VMAs satisfying specified subset of permissions. E.g., specifying +* PROCMAP_QUERY_VMA_READABLE only will return both readable and read/write VMAs, +* while having PROCMAP_QUERY_VMA_READABLE | PROCMAP_QUERY_VMA_WRITABLE will only +* return read/write VMAs, though both executable/non-executable and +* private/shared will be ignored. +* +* PROCMAP_QUERY_VMA_* flags are also returned in procmap_query.vma_flags +* field to specify actual VMA permissions. +*/ +PROCMAP_QUERY_VMA_READABLE =0x01 , +PROCMAP_QUERY_VMA_WRITABLE =0x02 , +PROCMAP_QUERY_VMA_EXECUTABLE =0x04 , +PROCMAP_QUERY_VMA_SHARED =0x08 , +/* +* Query modifier flags. +* +* By default VMA that covers provided address is returned, or -ENOENT +* is returned. With PROCMAP_QUERY_COVERING_OR_NEXT_VMA flag set, closest +* VMA with vma_start > addr will be returned if no covering VMA is +* found. +* +* PROCMAP_QUERY_FILE_BACKED_VMA instructs query to consider only VMAs that +* have file backing. Can be combined with PROCMAP_QUERY_COVERING_OR_NEXT_VMA +* to iterate all VMAs with file backing. +*/ +PROCMAP_QUERY_COVERING_OR_NEXT_VMA =0x10 , +PROCMAP_QUERY_FILE_BACKED_VMA =0x20 , +}; + +/* +* Input/output argument structured passed into ioctl() call. It can be used +* to query a set of VMAs (Virtual Memory Areas) of a process. +* +* Each field can be one of three kinds, marked in a short comment to the +* right of the field: +* - "in", input argument, user has to provide this value, kernel doesn't modify it; +* - "out", output argument, kernel sets this field with VMA data; +* - "in/out", input and output argument; user provides initial value (used +* to specify maximum allowable buffer size), and kernel sets it to actual +* amount of data written (or zero, if there is no data). +* +* If matching VMA is found (according to criterias specified by +* query_addr/query_flags, all the out fields are filled out, and ioctl() +* returns 0. If there is no matching VMA, -ENOENT will be returned. +* In case of any other error, negative error code other than -ENOENT is +* returned. +* +* Most of the data is similar to the one returned as text in /proc/<pid>/maps +* file, but procmap_query provides more querying flexibility. There are no +* consistency guarantees between subsequent ioctl() calls, but data returned +* for matched VMA is self-consistent. +*/ +struct procmap_query { +/* Query struct size, for backwards/forward compatibility */ +__u64 size; +/* +* Query flags, a combination of enum procmap_query_flags values. +* Defines query filtering and behavior, see enum procmap_query_flags. +* +* Input argument, provided by user. Kernel doesn't modify it. +*/ +__u64 query_flags;/* in */ +/* +* Query address. By default, VMA that covers this address will +* be looked up. PROCMAP_QUERY_* flags above modify this default +* behavior further. +* +* Input argument, provided by user. Kernel doesn't modify it. +*/ +__u64 query_addr;/* in */ +/* VMA starting (inclusive) and ending (exclusive) address, if VMA is found. */ +__u64 vma_start;/* out */ +__u64 vma_end;/* out */ +/* VMA permissions flags. A combination of PROCMAP_QUERY_VMA_* flags. */ +__u64 vma_flags;/* out */ +/* VMA backing page size granularity. */ +__u32 vma_page_size;/* out */ +/* +* VMA file offset. If VMA has file backing, this specifies offset +* within the file that VMA's start address corresponds to. +* Is set to zero if VMA has no backing file. +*/ +__u64 vma_offset;/* out */ +/* Backing file's inode number, or zero, if VMA has no backing file. */ +__u64 inode;/* out */ +/* Backing file's device major/minor number, or zero, if VMA has no backing file. */ +__u32 dev_major;/* out */ +__u32 dev_minor;/* out */ +/* +* If set to non-zero value, signals the request to return VMA name +* (i.e., VMA's backing file's absolute path, with " (deleted)" suffix +* appended, if file was unlinked from FS) for matched VMA. VMA name +* can also be some special name (e.g., "[heap]", "[stack]") or could +* be even user-supplied with prctl(PR_SET_VMA, PR_SET_VMA_ANON_NAME). +* +* Kernel will set this field to zero, if VMA has no associated name. +* Otherwise kernel will return actual amount of bytes filled in +* user-supplied buffer (see vma_name_addr field below), including the +* terminating zero. +* +* If VMA name is longer that user-supplied maximum buffer size, +* -E2BIG error is returned. +* +* If this field is set to non-zero value, vma_name_addr should point +* to valid user space memory buffer of at least vma_name_size bytes. +* If set to zero, vma_name_addr should be set to zero as well +*/ +__u32 vma_name_size;/* in/out */ +/* +* If set to non-zero value, signals the request to extract and return +* VMA's backing file's build ID, if the backing file is an ELF file +* and it contains embedded build ID. +* +* Kernel will set this field to zero, if VMA has no backing file, +* backing file is not an ELF file, or ELF file has no build ID +* embedded. +* +* Build ID is a binary value (not a string). Kernel will set +* build_id_size field to exact number of bytes used for build ID. +* If build ID is requested and present, but needs more bytes than +* user-supplied maximum buffer size (see build_id_addr field below), +* -E2BIG error will be returned. +* +* If this field is set to non-zero value, build_id_addr should point +* to valid user space memory buffer of at least build_id_size bytes. +* If set to zero, build_id_addr should be set to zero as well +*/ +__u32 build_id_size;/* in/out */ +/* +* User-supplied address of a buffer of at least vma_name_size bytes +* for kernel to fill with matched VMA's name (see vma_name_size field +* description above for details). +* +* Should be set to zero if VMA name should not be returned. +*/ +__u64 vma_name_addr;/* in */ +/* +* User-supplied address of a buffer of at least build_id_size bytes +* for kernel to fill with matched VMA's ELF build ID, if available +* (see build_id_size field description above for details). +* +* Should be set to zero if build ID should not be returned. +*/ +__u64 build_id_addr;/* in */ +}; + +#endif /* _UAPI_LINUX_FS_H */
Signature · 1 line
2.43.0
Andrii Nakryiko
maintainerPATCH v5 6/6 selftests/proc: add PROCMAP_QUERY ioctl tests
In reply to Andrii NakryikoExtend existing proc-pid-vm.c tests with PROCMAP_QUERY ioctl() API. Test a few successful and negative cases, validating querying filtering and exact vs next VMA logic works as expected.
--- tools/testing/selftests/proc/Makefile | 1 + tools/testing/selftests/proc/proc-pid-vm.c | 86 ++++++++++++++++++++++ 2 files changed, 87 insertions(+)
tools/testing/selftests/proc/Makefile +1 -0
@@ -1,6 +1,7 @@ # SPDX-License-Identifier: GPL-2.0-only CFLAGS += -Wall -O2 -Wno-unused-function CFLAGS += -D_GNU_SOURCE +CFLAGS += $(TOOLS_INCLUDES) LDFLAGS += -pthread TEST_GEN_PROGS :=
tools/testing/selftests/proc/proc-pid-vm.c +86 -0
98 lines not on this page · raw message
Signature · 1 line
2.43.0