| Message ID | 1625bc89782bf83d9d8c7c63e8ffcb651ccb15fa.1629655338.git.olivier@trillion01.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | coredump: io_uring: Cancel io_uring to avoid core truncation | expand |
diff --git a/fs/coredump.c b/fs/coredump.c index 07afb5ddb1c4..9aceb4b3b40d 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -41,6 +41,7 @@ #include <linux/fs.h> #include <linux/path.h> #include <linux/timekeeping.h> +#include <linux/io_uring.h> #include <linux/uaccess.h> #include <asm/mmu_context.h> @@ -625,6 +626,8 @@ void do_coredump(const kernel_siginfo_t *siginfo) need_suid_safe = true; } + io_uring_task_cancel(); + retval = coredump_wait(siginfo->si_signo, &core_state); if (retval < 0) goto fail_creds;
The previous solution of ignoring the TIF_NOTIFY_SIGNAL bit while dumping core is only working when the core dump is sent in a file. When a pipe is used, pipe_write returns -ERESTARTSYS if signal_pending which includes TIF_NOTIFY_SIGNAL is true. A more robust solution is to make sure that io_uring will not set TIF_NOTIFY_SIGNAL while the core dump is generated by cancelling all the io_uring requests made by the current task before starting. Fixes: 06af8679449d ("coredump: Limit what can interrupt coredumps") Signed-off-by: Olivier Langlois <olivier@trillion01.com> --- fs/coredump.c | 3 +++ 1 file changed, 3 insertions(+)