fs/nvs: Switch to use flash_flatten instead of flash_erase

The NVS currently requires explict erase capability of
a device to work, so usage of flash_erase has been replaced
with flash_flatten.
There has been additional LOG_WRN added to warn user that
NVS may not efficiently work with device that do not really
have erase.

Currently NVS relies on devices that require erase.

Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>

tests/subsys/fs/nvs/src/main.c

@@ -207,7 +207,18 @@ ZTEST_F(nvs, test_nvs_corrupted_write)
 		   &flash_max_write_calls);
 	stats_walk(fixture->sim_stats, flash_sim_write_calls_find, &flash_write_stat);
 
+#if defined(CONFIG_FLASH_SIMULATOR_EXPLICIT_ERASE)
 	*flash_max_write_calls = *flash_write_stat - 1;
+#else
+	/* When there is no explicit erase, erase is done with write, which means
+	 * that there are more writes needed. The nvs_write here will cause erase
+	 * to be called, which in turn calls the flash_fill; flash_fill will
+	 * overwrite data using buffer of size CONFIG_FLASH_FILL_BUFFER_SIZE,
+	 * and then two additional real writes are allowed.
+	 */
+	*flash_max_write_calls = (fixture->fs.sector_size /
+				  CONFIG_FLASH_FILL_BUFFER_SIZE) + 2;
+#endif
 	*flash_write_stat = 0;
 
 	/* Flash simulator will lose part of the data at the end of this write.

tests/subsys/fs/nvs/testcase.yaml

@@ -6,6 +6,9 @@ tests:
   filesystem.nvs.0x00:
     extra_args: DTC_OVERLAY_FILE=boards/qemu_x86_ev_0x00.overlay
     platform_allow: qemu_x86
+  filesystem.nvs.sim.no_erase:
+    extra_args: CONFIG_FLASH_SIMULATOR_EXPLICIT_ERASE=n
+    platform_allow: qemu_x86
   filesystem.nvs.cache:
     extra_args:
       - CONFIG_NVS_LOOKUP_CACHE=y