scsi: target: fix recursive locking in __configfs_open_file()
Subsystems (3)
PATCH v2 1/1 scsi: target: fix recursive locking in __configfs_open_file()
In flush_write_buffer, &p->frag_sem is acquired and then the loaded store function is called, which, here, is target_core_item_dbroot_store(). This function called filp_open(), following which these functions were called (in reverse order), according to the call trace:
down_read__configfs_open_filedo_dentry_openvfs_opendo_openpath_openatdo_filp_openfile_open_namefilp_opentarget_core_item_dbroot_storeflush_write_bufferconfigfs_write_iter
target_core_item_dbroot_store() tries to validate the new file path by trying to open the file path provided to it; however, in this case, the bug report shows:
db_root: not a directory: /sys/kernel/config/target/dbroot
indicating that the same configfs file was tried to be opened, on which it is currently working on. Thus, it is trying to acquire frag_sem semaphore of the same file of which it already holds the semaphore obtained in flush_write_buffer(), leading to acquiring the semaphore in a nested manner and a possibility of recursive locking.
Fix this by modifying target_core_item_dbroot_store() to use kern_path() instead of filp_open() to avoid opening the file using filesystem-specific function __configfs_open_file(), and further modifying it to make this fix compatible.
---
Changes since v1:
- Update commit message to reflect the fact that same file, which code was currently operating on, was tried to be opened again, leading to acquiring the same semaphore in nested manner & possibility of recursive locking.
v1 link: https://lore.kernel.org/all/[email protected]/T/
drivers/target/target_core_configfs.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)
drivers/target/target_core_configfs.c +7 -6
@@ -108,8 +108,8 @@ static ssize_t target_core_item_dbroot_store(struct config_item *item, const char *page, size_t count) { ssize_t read_bytes; -struct file *fp;ssize_t r = -EINVAL ; +struct path path = {};mutex_lock (&target_devices_lock);if (target_devices) {
@@ -131,17 +131,18 @@ static ssize_t target_core_item_dbroot_store(struct config_item *item, db_root_stage[read_bytes -1 ] ='\0' ;/* validate new db root before accepting it */ - fp =filp_open (db_root_stage,O_RDONLY ,0 ); -if (IS_ERR (fp)) { + r =kern_path (db_root_stage,LOOKUP_FOLLOW , &path); +if (r) {pr_err ("db_root: cannot open: %s\n" , db_root_stage); goto unlock; } -if (!S_ISDIR (file_inode (fp)->i_mode)) { -filp_close (fp,NULL ); +if (!d_is_dir (path.dentry)) { +path_put (&path);pr_err ("db_root: not a directory: %s\n" , db_root_stage); + r = -ENOTDIR ; goto unlock; } -filp_close (fp,NULL ); +path_put (&path);strscpy (db_root, db_root_stage);pr_debug ("Target_Core_ConfigFS: db_root set to %s\n" , db_root);
base-commit: 3a8660878839faadb4f1a6dd72c3179c1df56787
Signature · 1 line
2.34.1
Quoted 91 lines · Prithvi Tambewagh
In
flush_write_buffer,&p->frag_semis acquired and then the loaded store function is called, which, here, istarget_core_item_dbroot_store(). This function calledfilp_open(), following which these functions were called (in reverse order), according to the call trace:
down_read__configfs_open_filedo_dentry_openvfs_opendo_openpath_openatdo_filp_openfile_open_namefilp_opentarget_core_item_dbroot_storeflush_write_bufferconfigfs_write_iter
target_core_item_dbroot_store()tries to validate the new file path by trying to open the file path provided to it; however, in this case, the bug report shows:
db_root: not a directory:/sys/kernel/config/target/dbrootindicating that the same configfs file was tried to be opened, on which it is currently working on. Thus, it is trying to acquire
frag_semsemaphore of the same file of which it already holds the semaphore obtained inflush_write_buffer(), leading to acquiring the semaphore in a nested manner and a possibility of recursive locking.Fix this by modifying
target_core_item_dbroot_store()to usekern_path()instead offilp_open()to avoid opening the file using filesystem-specific function__configfs_open_file(), and further modifying it to make this fix compatible.Reported-by [email protected]Tested-by [email protected]Signed-off-by Prithvi Tambewagh <[email protected]>---
Changes since v1:
- Update commit message to reflect the fact that same file, which code was currently operating on, was tried to be opened again, leading to acquiring the same semaphore in nested manner & possibility of recursive locking.v1 link: https://lore.kernel.org/all/[email protected]/T/
drivers/target/target_core_configfs.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)drivers/target/target_core_configfs.c +7
-6@@ -108,8 +108,8 @@ static ssize_t target_core_item_dbroot_store(struct config_item *item, const char *page, size_t count) { ssize_t read_bytes; -struct file *fp;ssize_t r = -EINVAL ; +struct path path = {};mutex_lock (&target_devices_lock);if (target_devices) {@@ -131,17 +131,18 @@ static ssize_t target_core_item_dbroot_store(struct config_item *item, db_root_stage[read_bytes -1 ] ='\0' ;/* validate new db root before accepting it */ - fp =filp_open (db_root_stage,O_RDONLY ,0 ); -if (IS_ERR (fp)) { + r =kern_path (db_root_stage,LOOKUP_FOLLOW , &path); +if (r) {pr_err ("db_root: cannot open: %s\n" , db_root_stage); goto unlock; } -if (!S_ISDIR (file_inode (fp)->i_mode)) { -filp_close (fp,NULL ); +if (!d_is_dir (path.dentry)) { +path_put (&path);pr_err ("db_root: not a directory: %s\n" , db_root_stage); + r = -ENOTDIR ; goto unlock; } -filp_close (fp,NULL ); +path_put (&path);strscpy (db_root, db_root_stage);pr_debug ("Target_Core_ConfigFS: db_root set to %s\n" , db_root);
Quoted 4 lines
base-commit:
3a8660878839faadb4f1a6dd72c3179c1df56787Signature · 1 line
2.34.1
Hello all,
Just a gentle ping on this thread
Thanks,
Prithvi