ID: [88db46fd]

esp_new_jpeg: Fix some esp_new_jpeg bugs
1. Support IDF5.4
2. Support SD power supply comes from internal LDO IO in test_app
3. Add 16-byte alignment check for encoder input buffer when using ESP32-S3
4. Add 16-byte alignment check for decoder output buffer when using ESP32-S3
5. Fix inaccurate `inbuf_remain` member in `jpeg_dec_io_t` structure after decoding is finished
6. Fix spelling errors in test_app

Author: linxu

esp_new_jpeg/CHANGELOG.md

@@ -1,5 +1,19 @@
 # Changelog
 
+## v0.6.0
+
+### Features
+
+- Support IDF5.4
+- Support SD power supply comes from internal LDO IO in test_app
+- Add 16-byte alignment check for encoder input buffer when using ESP32-S3
+- Add 16-byte alignment check for decoder output buffer when using ESP32-S3
+
+### Bug Fixes
+
+- Fix inaccurate `inbuf_remain` member in `jpeg_dec_io_t` structure after decoding is finished
+- Fix spelling errors in test_app
+
 ## v0.5.1
 
 ### Bug Fixes

esp_new_jpeg/README.md

@@ -169,9 +169,9 @@ Please refer to the `test_app` folder for more details on API usage.
     >>> Image.open("file_name.jpg").info.get('progressive', 0)
     ```
 
-2. Why does the decoded output image appear misaligned?
+2. Why does the output image appear misaligned?
 
-   The issue typically occurs when a few columns of the image on the far left or right side appear on the opposite side. If you are using the ESP32-S3, a possible cause is that the output buffer is not 16-byte aligned. Please use the `jpeg_calloc_align` function to allocate output buffer.
+   The issue typically occurs when a few columns of the image on the far left or right side appear on the opposite side. If you are using the ESP32-S3, a possible cause is that the output buffer of decoder or the input buffer of encoder is not 16-byte aligned. Please use the `jpeg_calloc_align` function to allocate buffer.
 
 3. How can I use a simpler method to check if the encoder's output is correct?
 
@@ -188,7 +188,7 @@ Please refer to the `test_app` folder for more details on API usage.
 
 The following table shows the support of ESP_NEW_JPEG for Espressif SoCs. The "✔" means supported, and the "✖" means not supported.
 
-| Chip     | v0.5.1   |
+| Chip     | v0.6.0   |
 |----------|----------|
 | ESP32    | ✔ |
 | ESP32-S2 | ✔ |

esp_new_jpeg/idf_component.yml

@@ -1,4 +1,4 @@
-version: "0.5.1"
+version: "0.6.0"
 description: Espressif JPEG encoder and decoder
 url: https://github.com/espressif/esp-adf-libs/tree/master/esp_new_jpeg
 repository: https://github.com/espressif/esp-adf-libs.git

esp_new_jpeg/include/esp_jpeg_dec.h

@@ -137,6 +137,8 @@ jpeg_error_t jpeg_dec_get_process_count(jpeg_dec_handle_t jpeg_dec, int *process
 /**
  * @brief  Decode JPEG picture
  *
+ * @note  For ESP32-S3, please use `jpeg_calloc_align` to allocate a 16-byte aligned output buffer.
+ *
  * @param[in]  jpeg_dec  JPEG decoder handle
  * @param[in]  io        Pointer to io control struct
  *

esp_new_jpeg/include/esp_jpeg_enc.h

@@ -58,6 +58,8 @@ jpeg_error_t jpeg_enc_open(jpeg_enc_config_t *info, jpeg_enc_handle_t *jpeg_enc)
 /**
  * @brief  Encode one image
  *
+ * @note  For ESP32-S3, please use `jpeg_calloc_align` to allocate a 16-byte aligned input buffer.
+ *
  * @param[in]   jpeg_enc     The JPEG encoder handle. It gained from `jpeg_enc_open`
  * @param[in]   in_buf       The input buffer, It needs a completed image. The buffer must be aligned 16 byte.
  * @param[in]   inbuf_size   The size of `in_buf`. The value must be size of a completed image.

esp_new_jpeg/include/esp_jpeg_version.h

@@ -7,7 +7,7 @@
 extern "C" {
 #endif  /* __cplusplus */
 
-#define ESP_JPEG_VERION "0.5.1"
+#define ESP_JPEG_VERION "0.6.0"
 
 /**
  * @version 0.5.0: