Hi Jeff,
Could you fix this in your vfs_inode_has_locks patch ?
We should add one dummy inline func in :
#else /* !CONFIG_FILE_LOCKING */
#endif
Currently I just added your filelock patch as one [DO NOT MERGE] in ceph
tree for testing.
Thanks!
- Xiubo
On 30/11/2022 14:39, kernel test robot wrote:
tree: https://github.com/ceph/ceph-client.git testing
head: 72ead199864c6617e424df714e73fde562504837
commit: 527fba16ba73a44d115dc0521ad756c68a0ab1f6 [2/4] ceph: switch to vfs_inode_has_locks() to fix file lock bug
config: ia64-randconfig-r015-20221130
compiler: ia64-linux-gcc (GCC) 12.1.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://github.com/ceph/ceph-client/commit/527fba16ba73a44d115dc0521ad756c68a0ab1f6
git remote add ceph-client https://github.com/ceph/ceph-client.git
git fetch --no-tags ceph-client testing
git checkout 527fba16ba73a44d115dc0521ad756c68a0ab1f6
# save the config file
mkdir build_dir && cp config build_dir/.config
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-12.1.0 make.cross W=1 O=build_dir ARCH=ia64 SHELL=/bin/bash
If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@xxxxxxxxx>
All errors (new ones prefixed by >>):
fs/ceph/caps.c: In function 'ceph_get_caps':
fs/ceph/caps.c:2967:21: error: implicit declaration of function 'vfs_inode_has_locks' [-Werror=implicit-function-declaration]
2967 | if (vfs_inode_has_locks(inode))
| ^~~~~~~~~~~~~~~~~~~
cc1: some warnings being treated as errors
vim +/vfs_inode_has_locks +2967 fs/ceph/caps.c
2941
2942 /*
2943 * Wait for caps, and take cap references. If we can't get a WR cap
2944 * due to a small max_size, make sure we check_max_size (and possibly
2945 * ask the mds) so we don't get hung up indefinitely.
2946 */
2947 int ceph_get_caps(struct file *filp, int need, int want, loff_t endoff, int *got)
2948 {
2949 struct ceph_file_info *fi = filp->private_data;
2950 struct inode *inode = file_inode(filp);
2951 struct ceph_inode_info *ci = ceph_inode(inode);
2952 struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
2953 int ret, _got, flags;
2954
2955 ret = ceph_pool_perm_check(inode, need);
2956 if (ret < 0)
2957 return ret;
2958
2959 if ((fi->fmode & CEPH_FILE_MODE_WR) &&
2960 fi->filp_gen != READ_ONCE(fsc->filp_gen))
2961 return -EBADF;
2962
2963 flags = get_used_fmode(need | want);
2964
2965 while (true) {
2966 flags &= CEPH_FILE_MODE_MASK;
2967 if (vfs_inode_has_locks(inode))
2968 flags |= CHECK_FILELOCK;
2969 _got = 0;
2970 ret = try_get_cap_refs(inode, need, want, endoff,
2971 flags, &_got);
2972 WARN_ON_ONCE(ret == -EAGAIN);
2973 if (!ret) {
2974 struct ceph_mds_client *mdsc = fsc->mdsc;
2975 struct cap_wait cw;
2976 DEFINE_WAIT_FUNC(wait, woken_wake_function);
2977
2978 cw.ino = ceph_ino(inode);
2979 cw.tgid = current->tgid;
2980 cw.need = need;
2981 cw.want = want;
2982
2983 spin_lock(&mdsc->caps_list_lock);
2984 list_add(&cw.list, &mdsc->cap_wait_list);
2985 spin_unlock(&mdsc->caps_list_lock);
2986
2987 /* make sure used fmode not timeout */
2988 ceph_get_fmode(ci, flags, FMODE_WAIT_BIAS);
2989 add_wait_queue(&ci->i_cap_wq, &wait);
2990
2991 flags |= NON_BLOCKING;
2992 while (!(ret = try_get_cap_refs(inode, need, want,
2993 endoff, flags, &_got))) {
2994 if (signal_pending(current)) {
2995 ret = -ERESTARTSYS;
2996 break;
2997 }
2998 wait_woken(&wait, TASK_INTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT);
2999 }
3000
3001 remove_wait_queue(&ci->i_cap_wq, &wait);
3002 ceph_put_fmode(ci, flags, FMODE_WAIT_BIAS);
3003
3004 spin_lock(&mdsc->caps_list_lock);
3005 list_del(&cw.list);
3006 spin_unlock(&mdsc->caps_list_lock);
3007
3008 if (ret == -EAGAIN)
3009 continue;
3010 }
3011
3012 if ((fi->fmode & CEPH_FILE_MODE_WR) &&
3013 fi->filp_gen != READ_ONCE(fsc->filp_gen)) {
3014 if (ret >= 0 && _got)
3015 ceph_put_cap_refs(ci, _got);
3016 return -EBADF;
3017 }
3018
3019 if (ret < 0) {
3020 if (ret == -EFBIG || ret == -EUCLEAN) {
3021 int ret2 = ceph_wait_on_async_create(inode);
3022 if (ret2 < 0)
3023 return ret2;
3024 }
3025 if (ret == -EFBIG) {
3026 check_max_size(inode, endoff);
3027 continue;
3028 }
3029 if (ret == -EUCLEAN) {
3030 /* session was killed, try renew caps */
3031 ret = ceph_renew_caps(inode, flags);
3032 if (ret == 0)
3033 continue;
3034 }
3035 return ret;
3036 }
3037
3038 if (S_ISREG(ci->netfs.inode.i_mode) &&
3039 ceph_has_inline_data(ci) &&
3040 (_got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) &&
3041 i_size_read(inode) > 0) {
3042 struct page *page =
3043 find_get_page(inode->i_mapping, 0);
3044 if (page) {
3045 bool uptodate = PageUptodate(page);
3046
3047 put_page(page);
3048 if (uptodate)
3049 break;
3050 }
3051 /*
3052 * drop cap refs first because getattr while
3053 * holding * caps refs can cause deadlock.
3054 */
3055 ceph_put_cap_refs(ci, _got);
3056 _got = 0;
3057
3058 /*
3059 * getattr request will bring inline data into
3060 * page cache
3061 */
3062 ret = __ceph_do_getattr(inode, NULL,
3063 CEPH_STAT_CAP_INLINE_DATA,
3064 true);
3065 if (ret < 0)
3066 return ret;
3067 continue;
3068 }
3069 break;
3070 }
3071 *got = _got;
3072 return 0;
3073 }
3074
