Clean & add logs for error condition [Description] Clean [Module] kernel [Board]...

diff --git a/os/fs/driver/mtd/smart.c b/os/fs/driver/mtd/smart.c
index 2f246bc..91e7285 100644 (file)
--- a/os/fs/driver/mtd/smart.c
+++ b/os/fs/driver/mtd/smart.c
@@ -1948,7 +1948,6 @@ static int smart_scan(FAR struct smart_struct_s *dev)
                                        }
                                }                               //if(logicalsector == SMART_BAD_SECTOR_NUMBER)
 #endif                                                 /*CONFIG_SMARTFS_BAD_SECTOR */
-
                        } else {
                                fvdbg("logicalsector : physicalsector -> %d : %d; status_released: %d, status_committed: %d\n", logicalsector, sector, status_released, status_committed);
                        }
@@ -4173,7 +4172,7 @@ relocate_good_sector:
                oldphyssector = physsector;
                physsector = smart_findfreephyssector(dev, FALSE);
                if (physsector == 0xFFFF) {
-                       //fdbg("Error relocating sector %d\n", req->logsector);
+                       fdbg("Error relocating sector %d\n", req->logsector);
                        ret = -EIO;
                        goto errout;
                }
@@ -5435,7 +5434,7 @@ int smart_recoversectors(FAR struct inode *inode, char *validsectors, int *nobso
                status_committed = SECTOR_IS_COMMITTED(header);
 
                if (logicalsector > 0 && status_committed && !status_released && !GET_VAL(validsectors, sector)) {
-                       fdbg("RECOVERY: sector %u CORRUPTED!!\n", sector);
+                       fdbg("RECOVERY: logical %d phy : %d sector %u CORRUPTED!!\n", logicalsector, physsector, sector);
                        (*nobsolete)++;
 
 #if CONFIG_SMARTFS_ERASEDSTATE == 0xFF

diff --git a/os/fs/smartfs/smartfs_smart.c b/os/fs/smartfs/smartfs_smart.c
index 5fb6be5..fa641c1 100644 (file)
--- a/os/fs/smartfs/smartfs_smart.c
+++ b/os/fs/smartfs/smartfs_smart.c
@@ -978,7 +978,7 @@ static ssize_t smartfs_write(FAR struct file *filep, const char *buffer, size_t
                                readwrite.count = sizeof(uint16_t);
 #ifdef CONFIG_SMARTFS_JOURNALING
                                ret = smartfs_create_journalentry(fs, T_WRITE, readwrite.logsector,
-                                       readwrite.offset, readwrite.count, 0, 0, readwrite.buffer, &t_sector, &t_offset);
+                                                                                                 readwrite.offset, readwrite.count, 0, 0, readwrite.buffer, &t_sector, &t_offset);
                                if (ret != OK) {
                                        fdbg("Journal entry creation failed.\n");
                                        goto errout_with_semaphore;

diff --git a/os/fs/smartfs/smartfs_utils.c b/os/fs/smartfs/smartfs_utils.c
index f1f4844..cfa0942 100644 (file)
--- a/os/fs/smartfs/smartfs_utils.c
+++ b/os/fs/smartfs/smartfs_utils.c
@@ -112,16 +112,16 @@ extern size_t smart_sect_header_size;
 #endif
 
 #ifdef CONFIG_SMARTFS_ALIGNED_ACCESS
-#define ENTRY_VALID(e) ((smartfs_rdle16(&e->flags) & SMARTFS_DIRENT_EMPTY) != \
-                                               (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_EMPTY)) && \
-                                               ((smartfs_rdle16(&e->flags) & SMARTFS_DIRENT_ACTIVE) == \
-                                               (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_ACTIVE))
+#define ENTRY_VALID(e) ((smartfs_rdle16(&(e)->flags) & SMARTFS_DIRENT_EMPTY) != \
+                        (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_EMPTY)) && \
+                        ((smartfs_rdle16(&(e)->flags) & SMARTFS_DIRENT_ACTIVE) == \
+                        (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_ACTIVE))
 
 #else
-#define ENTRY_VALID(e) ((e->flags & SMARTFS_DIRENT_EMPTY) != \
-                                               (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_EMPTY)) && \
-                                               ((e->flags & SMARTFS_DIRENT_ACTIVE) == \
-                                               (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_ACTIVE))
+#define ENTRY_VALID(e) (((e)->flags & SMARTFS_DIRENT_EMPTY) != \
+                        (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_EMPTY)) && \
+                        (((e)->flags & SMARTFS_DIRENT_ACTIVE) == \
+                        (SMARTFS_ERASEDSTATE_16BIT & SMARTFS_DIRENT_ACTIVE))
 
 #endif
 
@@ -1274,6 +1274,7 @@ int smartfs_createentry(struct smartfs_mountpt_s *fs, uint16_t parentdirsector,
        readwrite.buffer = (uint8_t *)&fs->fs_rwbuffer[offset];
        ret = FS_IOCTL(fs, BIOC_WRITESECT, (unsigned long)&readwrite);
        if (ret < 0) {
+               fdbg("failed to write new entry to parent directory psector : %d\n", psector);
                goto errout;
        }
 
@@ -1945,6 +1946,9 @@ int smartfs_journal_init(struct smartfs_mountpt_s *fs)
        memset(journal->active_sectors, 0, mapsize);
 
        if (journal->jarea != -1) {
+#ifdef CONFIG_DEBUG_FS
+               print_journal_sectors(fs);
+#endif
                startsector = SMARTFS_LOGGING_SECTOR + journal->jarea * CONFIG_SMARTFS_NLOGGING_SECTORS;
                journal->sector = startsector;
                readsect = startsector;
@@ -2028,7 +2032,11 @@ int smartfs_journal_init(struct smartfs_mountpt_s *fs)
        }
        journal->sector = SMARTFS_LOGGING_SECTOR;
        journal->offset = 1;
-       return set_area_id_bits(fs, AREA_USED);
+       ret = set_area_id_bits(fs, AREA_USED);
+       if (ret != OK) {
+               fdbg("set_area_id_bits failed... bits : %d\n", AREA_USED);
+       }
+       return ret;
 
 err_out:
        /* Free in case of failure */
@@ -2126,7 +2134,7 @@ static int smartfs_get_journal_area(struct smartfs_mountpt_s *fs)
        return result[index[0]][index[1]];
 }
 
-#ifdef CONFIG_SMARTFS_JOURNALING_TEST
+#ifdef CONFIG_DEBUG_FS
 int print_journal_sectors(struct smartfs_mountpt_s *fs)
 {
        int ret;
@@ -2137,7 +2145,7 @@ int print_journal_sectors(struct smartfs_mountpt_s *fs)
        struct journal_transaction_manager_s *j_mgr;
        struct smartfs_logging_entry_s *entry;
 
-       fvdbg("Print all Journal sectors\n");
+       fdbg("Print all Journal sectors\n");
        j_mgr = fs->journal;
 
        if (!(j_mgr->enabled)) {
@@ -2163,22 +2171,29 @@ int print_journal_sectors(struct smartfs_mountpt_s *fs)
                                                break;
                                        }
                                }
-                               fdbg("*****Entry*****\n");
-                               fdbg("Transaction Type: %u\n", GET_TRANS_TYPE(entry->trans_info));
-                               fdbg("Transaction Started: %s\n", T_START_CHECK(entry->trans_info) ? "yes" : "no");
-                               fdbg("Transaction Finished: %s\n", T_FINISH_CHECK(entry->trans_info) ? "yes" : "no");
-                               fdbg("Transaction required a sync: %s\n", T_NEEDSYNC_CHECK(entry->trans_info) ? "yes" : "no");
-                               fdbg("sector: %u\n", entry->curr_sector);
-                               fdbg("offset: %u\n", entry->offset);
-                               fdbg("seq_no: %u\n", entry->seq_no);
-                               fdbg("datalen: %u\n", entry->datalen);
-                               fdbg("generic_1 %u\n", entry->generic_1);
-                               if (entry->datalen > 0 && GET_TRANS_TYPE(entry->trans_info) != T_DELETE) {
-                                       fdbg("Data: ");
-                                       for (i = 0; i < entry->datalen; i++) {
-                                               fsdbg("%02x ", j_mgr->buffer[sizeof(struct smartfs_logging_entry_s) + i]);
+                               if (!T_FINISH_CHECK(entry->trans_info)) {
+                                       fdbg("*****Entry*****\n");
+                                       fdbg("Transaction Type: %u\n", GET_TRANS_TYPE(entry->trans_info));
+                                       fdbg("Transaction Started: %s\n", T_START_CHECK(entry->trans_info) ? "yes" : "no");
+                                       fdbg("Transaction Finished: %s\n", T_FINISH_CHECK(entry->trans_info) ? "yes" : "no");
+                                       fdbg("Transaction required a sync: %s\n", T_NEEDSYNC_CHECK(entry->trans_info) ? "yes" : "no");
+                                       fdbg("sector: %u\n", entry->curr_sector);
+                                       fdbg("offset: %u\n", entry->offset);
+                                       fdbg("seq_no: %u\n", entry->seq_no);
+                                       fdbg("datalen: %u\n", entry->datalen);
+                                       fdbg("generic_1 %u\n", entry->generic_1);
+#ifdef CONFIG_DEBUG_FS
+#ifndef DEBUG_VERBOSE
+                                       if (entry->datalen > 0 && GET_TRANS_TYPE(entry->trans_info) != T_DELETE)
+#endif
+                                       {
+                                               fdbg("Data: ");
+                                               for (i = 0; i < entry->datalen; i++) {
+                                                       fsdbg("%02x ", j_mgr->buffer[sizeof(struct smartfs_logging_entry_s) + i]);
+                                               }
+                                               fsdbg("\n");
                                        }
-                                       fsdbg("\n");
+#endif
                                }
                        } else {
                                break;
@@ -2532,6 +2547,9 @@ static int smartfs_redo_rename(struct smartfs_mountpt_s *fs)
                req.count = sizeof(direntry->flags);
                req.buffer = (uint8_t *)&direntry->flags;
                ret = FS_IOCTL(fs, BIOC_WRITESECT, (unsigned long)&req);
+               if (ret != OK) {
+                       fdbg("Inactive old entry failed... offset : %d\n", req.offset);
+               }
        } else {
                ret = OK;
        }
@@ -2711,7 +2729,7 @@ static int process_transaction(struct smartfs_mountpt_s *fs)
        /* Check whether the transaction qualifies for redo */
        if (T_START_CHECK(entry->trans_info) && !T_FINISH_CHECK(entry->trans_info)) {
                /* Choose redo routine based on transaction type */
-               fvdbg("type = %d\n", GET_TRANS_TYPE(entry->trans_info));
+               fdbg("type : %d sector : %d generic_1 : %d\n", GET_TRANS_TYPE(entry->trans_info), entry->curr_sector, entry->generic_1);
                switch (GET_TRANS_TYPE(entry->trans_info)) {
                case T_SYNC:
                case T_WRITE:
@@ -3067,6 +3085,7 @@ static int smartfs_set_sequence(struct smartfs_mountpt_s *fs, uint16_t sector, u
        struct smart_read_write_s req;
        struct smartfs_logging_entry_s *entry;
        struct journal_transaction_manager_s *j_mgr;
+       int ret;
 
        j_mgr = fs->journal;
        entry = (struct smartfs_logging_entry_s *)(j_mgr->buffer);
@@ -3077,8 +3096,12 @@ static int smartfs_set_sequence(struct smartfs_mountpt_s *fs, uint16_t sector, u
 
        req.count = sizeof(entry->seq_no);
        req.buffer = (uint8_t *)&(entry->seq_no);
-       if ((req.offset + req.count <= j_mgr->availbytes) && (FS_IOCTL(fs, BIOC_WRITESECT, (unsigned long)&req) == OK)) {
-               return OK;
+       if (req.offset + req.count <= j_mgr->availbytes)  {
+               ret = FS_IOCTL(fs, BIOC_WRITESECT, (unsigned long)&req);
+               if (ret != OK) {
+                       fdbg("set sequence failed.. seq no : %d logicsector : %d\n", seq, sector);
+               }
+               return ret;
        }
        return ERROR;
 }
@@ -3211,7 +3234,7 @@ static int remove_from_list(struct journal_transaction_manager_s *j_mgr, uint16_
        }
        SET_INACTIVE(j_mgr->active_sectors, sector);
 errout:
-       return ret;;
+       return ret;
 }
 
 /****************************************************************************
@@ -3291,6 +3314,7 @@ int smartfs_create_journalentry(struct smartfs_mountpt_s *fs, enum logging_trans
                }
                break;
        }
+       fvdbg("Create journalentry : %d\n", *t_sector);
        return OK;
 }