Qualcomm笔记

最新推荐文章于 2022-11-10 20:55:20 发布
最新推荐文章于 2022-11-10 20:55:20 发布
阅读量8.3k 收藏 5
点赞数 1
分类专栏： Qualcomm
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/TopicDev/article/details/18262955
版权

[1]RAM跑什么频率

cat  /sys/kernel/debug/clk/bimc_clk/measure

[2]

root@c1_dcg:/sys/class/leds/lcd-backlight # cat brightness
cat brightness

 

[3]root@c1_dcg:/ # cat proc/kmsg

[4]monkey --ignore-crashes --ignore-timeouts --kill-process-after-error --ignore-security-exceptions --throttle 1500 -v -v -v -s 5 60000000

 

[5]中断唤醒：
在终端输入下列命令：
# echo 1 > /sys/module/msm_show_resume_irq/parameters/debug_mask

[6]三十万前摄卡死，发现是mipi振幅太小，导致驱动能力不足，通过调整gc0339_lib.c的0x6a寄存器的振幅值与preview.h中的ae值改善

[7]

GPB1，即定时器1.

   定时器的输入频率 = PCLK /{prescaler value + 1} / {divider value}。

[8]

蓝牙鼠标在Setting界面边缘地带显示异常
此问题是高通平台问题，原因是SurfaceFlinger图层合成有问题

 Region::const_iterator head = layer->visibleRegion.begin(); 
 Region::const_iterator tail = layer->visibleRegion.end(); 
 while (head != tail) { 
 if (head->left <0 || head->top <0 || head->right >= hw->getWidth() |
 | head->bottom >= hw->getHeight() ) 
 { 
  cur->setSkip(true); 
  break; 
 } 
  head++; 
 }

SurfaceFlinger参考：

http://blog.csdn.net/luoshengyang/article/details/8079456

 

[9]

diff --git a/drivers/media/platform/msm/camera_v2/sensor/msm_sensor.c b/drivers/media/platform/msm/camera_v2/sensor/msm_sensor.c
index 0143a81..6bafaf2 100755
--- a/drivers/media/platform/msm/camera_v2/sensor/msm_sensor.c
+++ b/drivers/media/platform/msm/camera_v2/sensor/msm_sensor.c
@@ -1125,6 +1125,70 @@ static struct msm_camera_i2c_fn_t msm_sensor_qup_func_tbl = {
  .i2c_write_conf_tbl = msm_camera_qup_i2c_write_conf_tbl,
 };
 
+
+/* add sensor info

+*/
+static struct kobject *msm_sensor_device;
+static char module_info[80] = {0};
+
+static void msm_sensor_set_module_info(struct msm_sensor_ctrl_t *s_ctrl)
+{
+ printk("s_ctrl->sensordata->camera_type = %d\n", s_ctrl->sensordata->sensor_info->position);
+ switch (s_ctrl->sensordata->sensor_info->position) {
+ case BACK_CAMERA_B:
+  strcat(module_info, "back: ");
+  break;
+ case FRONT_CAMERA_B:
+  strcat(module_info, "front: ");
+  break;
+ default:
+  strcat(module_info, "unknown: ");
+  break;
+ }
+
+ strcat(module_info, s_ctrl->sensordata->sensor_name);
+ strcat(module_info, "\n");
+}
+
+static ssize_t msm_sensor_module_id_show(struct device *dev,
+  struct device_attribute *attr, char *buf)
+{
+ ssize_t rc = 0;
+
+ sprintf(buf, "%s\n", module_info);
+ rc = strlen(buf) + 1;
+
+ return rc;
+}
+
+static DEVICE_ATTR(sensor, 0444, msm_sensor_module_id_show, NULL);
+
+static int32_t msm_sensor_init_device_name(void)
+{
+ int32_t rc = 0;
+
+ msm_sensor_device = kobject_create_and_add("android_camera", NULL);
+ if (msm_sensor_device == NULL) {
+  pr_info("%s: subsystem_register failed\n", __func__);
+  rc = -ENOMEM;
+  return rc ;
+ }
+ rc = sysfs_create_file(msm_sensor_device, &dev_attr_sensor.attr);
+ if (rc) {
+  pr_info("%s: sysfs_create_file failed\n", __func__);
+  kobject_del(msm_sensor_device);
+ }
+
+ return 0 ;
+}
+/* add sensor info

+*/
+
+
 int32_t msm_sensor_platform_probe(struct platform_device *pdev, void *data)
 {
  int rc = 0;
@@ -1214,6 +1278,9 @@ int32_t msm_sensor_platform_probe(struct platform_device *pdev, void *data)
  msm_sd_register(&s_ctrl->msm_sd);
  CDBG("%s:%d\n", __func__, __LINE__);
 
+ msm_sensor_init_device_name();
+ msm_sensor_set_module_info(s_ctrl);
+
  s_ctrl->func_tbl->sensor_power_down(s_ctrl);
  CDBG("%s:%d\n", __func__, __LINE__);
  return rc;

 [10]monkey测试过程中遇到手机切到download模式的情况。从dump内核log看发生了data abort

日志里面有很多用户态的错误，一般这个不会导致重启。

重启原因是：
<3>[223906.016015] ion_sg_table: invalid handle passed to map_dma.
<4>[223906.020751] ------------[ cut here ]------------
<4>[223906.025369] WARNING: at /home/ubuntu/jenkins/jenkins/workspace/Build-W5256HTC-EU-8610-ARM11/label/SCM_glsrv4_Lin/kernel/drivers/gpu/ion/msm/ion_iommu_map.c:522 ion_unmap_iommu+0x12c/0x1ac()
<6>[223906.042347] ion_unmap_iommu: (5,0) was never mapped for 000bdd83
<6>[223906.042366] Modules linked in: [last unloaded: wlan]
<6>[223906.042424] [<c010cc94>] (unwind_backtrace+0x0/0x138) from [<c018b6a8>] (warn_slowpath_common+0x4c/0x64)
<6>[223906.042461] [<c018b6a8>] (warn_slowpath_common+0x4c/0x64) from [<c018b754>] (warn_slowpath_fmt+0x30/0x40)
<6>[223906.042498] [<c018b754>] (warn_slowpath_fmt+0x30/0x40) from [<c0478b60>] (ion_unmap_iommu+0x12c/0x1ac)
<6>[223906.042543] [<c0478b60>] (ion_unmap_iommu+0x12c/0x1ac) from [<c05c96fc>] (msm_isp_buf_unprepare+0xc8/0x120)
<6>[223906.042584] [<c05c96fc>] (msm_isp_buf_unprepare+0xc8/0x120) from [<c05c97a4>] (msm_isp_release_bufq+0x50/0x70)
<6>[223906.042622] [<c05c97a4>] (msm_isp_release_bufq+0x50/0x70) from [<c05ca290>] (msm_isp_proc_buf_cmd+0x7c/0x84)
<6>[223906.042662] [<c05ca290>] (msm_isp_proc_buf_cmd+0x7c/0x84) from [<c05cc0e0>] (msm_isp_ioctl+0x9c/0x354)
<6>[223906.042701] [<c05cc0e0>] (msm_isp_ioctl+0x9c/0x354) from [<c05e9a7c>] (subdev_do_ioctl+0x144/0x558)
<6>[223906.042741] [<c05e9a7c>] (subdev_do_ioctl+0x144/0x558) from [<c05dee78>] (video_usercopy+0x148/0x57c)
<6>[223906.042779] [<c05dee78>] (video_usercopy+0x148/0x57c) from [<c05de168>] (v4l2_ioctl+0x94/0x138)
<6>[223906.042817] [<c05de168>] (v4l2_ioctl+0x94/0x138) from [<c026ed48>] (do_vfs_ioctl+0x84/0x5ac)
<6>[223906.042852] [<c026ed48>] (do_vfs_ioctl+0x84/0x5ac) from [<c026f2dc>] (sys_ioctl+0x6c/0x7c)
<6>[223906.042887] [<c026f2dc>] (sys_ioctl+0x6c/0x7c) from [<c0106720>] (ret_fast_syscall+0x0/0x30)
<4>[223906.042910] ---[ end trace c716865583a4b41b ]---
<1>[223906.047027] Unable to handle kernel paging request at virtual address 000bdd87
<1>[223906.054252] pgd = f2f4c000
<1>[223906.057082] [000bdd87] *pgd=00000000
<0>[223906.060681] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
<6>[223906.066062] Modules linked in: [last unloaded: wlan]
<6>[223906.071099] CPU: 2 Tainted: G W O (3.4.0-g80ec9c8-01181-g969bc63-dirty #1)
<6>[223906.079089] PC is at ion_free+0x8/0xac
<6>[223906.082907] LR is at msm_isp_buf_unprepare+0xd8/0x120

这个看上去和显示有关

 

[11]

貌似是CPU接屏ID的管脚被屏损坏了。发现屏的这个ID脚在屏上是直接接到VDDIO上，没有串联电阻，我们主板上对应的GPIO口也没有串入限流的电阻。所以，高度怀疑是这个GPIO口在被直接接到电源VDDIO后造成损坏。

[12]

在Pre-CS2.1.21.3基线，YUV模组显示黑屏,RAW显示正常。

sensor.c,isp_stats.h,isp_hw.c，cpp_module.c

D/QCamera2HWI( 206): int qcamera::QCamera2HardwareInterface::startPreview(): X
D/QCamera2HWI( 206): [KPI Perf] static int qcamera::QCamera2HardwareInterface::start_preview(camera_device*): X
E/QCameraParameters( 206): [KPI Perf] int32_t qcamera::QCameraParameters::setFaceDetection(bool): PROFILE_FACE_DETECTION_VALUE = 1 num_fd = 5
D/QCameraParameters( 206): int32_t qcamera::QCameraParameters::initBatchUpdate(void*):Initializing batch parameter set
D/QCameraParameters( 206): int32_t qcamera::QCameraParameters::AddSetParmEntryToBatch(void*, cam_intf_parm_type_t, uint32_t, void*): num_entry: 1, paramType: 30, paramLength: 8, aligned_size_req: 24
E/mm-camera( 225): mct_pipeline_process_set:command=800000e
E/mm-camera( 225): mct_pipeline_send_ctrl_events: Send Set Parm events
D/QCameraParameters( 206): int32_t qcamera::QCameraParameters::commitSetBatch():waiting for commitSetBatch to complete
D/QCameraParameters( 206): int32_t qcamera::QCameraParameters::setFaceDetection(bool): FaceProcMask -> 1
D/Camera ( 2281): app passed NULL surface
I/CameraClient( 206): stopPreview: Disable picture related messages
E/QCamera2HWI( 206): [KPI Perf] static void qcamera::QCamera2HardwareInterface::stop_preview(camera_device*): E PROFILE_STOP_PREVIEW
D/QCamera2HWI( 206): int qcamera::QCamera2HardwareInterface::stopPreview(): E

Via above log:
the walkflow is :
start_preview->QCameraParameters::setFaceDetection-> initBatchUpdate-> AddSetParmEntryToBatch ->D/Camera ( 2281): app passed NULL surface

It seems that the is a Null surface passed from upper layer.
In normal preview, the log: "app passed Null surface doesn't appear".

 

catch yuv data.
static void isp_ch_util_dump_frame(int ion_fd, mct_stream_info_t *stream_info,
isp_frame_buffer_t *image_buf, uint32_t frame_idx)
{
...
CDBG("%s:frame_idx %d enabled %d streamtype %d width %d height %d",....

property_get("persist.camera.isp.dump", value, "0");
enabled = atoi(value);
if(!enabled){
sDumpFrmCnt = 0;
return;
}
if(sDumpFrmCnt >= 0 && sDumpFrmCnt <= frm_num) {
...
"/data/isp_dump_%d_preview_%d_%d.yuv"
// "/data/isp_dump_%d_video_%d_%d.yuv"
"/data/isp_dump_%d_postview_%d_%d.yuv"
"/data/isp_dump_%d_snapshot_%d_%d.yuv"

the yuv data will be saved in /data/
please try to use below cmd and upload yuv data.
adb shell setprop persist.camera.isp.dump 1

adb shell setprop persist.camera.isp.dump 1

取出preview, postview, snapshot的yuv data.
setprop persist.camera.isp.dump 1



 

diff --git a/QCamera2/HAL/QCamera2HWICallbacks.cpp b/QCamera2/HAL/QCamera2HWICallbacks.cpp
index 6b40ec3..a3f1614 100644
--- a/ QCamera2/HAL/QCamera2HWICallbacks.cpp
+++ b/ QCamera2/HAL/QCamera2HWICallbacks.cpp
@@ -1430,8 +1430,11 @@ void QCamera2HardwareInterface::dumpFrameToFile(QCameraStream *stream,
case QCAMERA_DUMP_FRM_SNAPSHOT:
{
mParameters.getStreamDimension(CAM_STREAM_TYPE_SNAPSHOT, dim);
- snprintf(buf, sizeof(buf), "/data/%ds_%dx%d_%d.yuv",
- mDumpFrmCnt, dim.width, dim.height, frame->frame_idx);
+ snprintf(buf, sizeof(buf), "%ds_%dx%d_%d.yuv",
+ mDumpFrmCnt,
+ dim.width,
+ dim.height,
+ frame->frame_idx);
}
break;
case QCAMERA_DUMP_FRM_VIDEO:
@@ -1443,15 +1446,21 @@ void QCamera2HardwareInterface::dumpFrameToFile(QCameraStream *stream,
case QCAMERA_DUMP_FRM_RAW:
{
mParameters.getStreamDimension(CAM_STREAM_TYPE_RAW, dim);
- snprintf(buf, sizeof(buf), "/data/%dr_%dx%d_%d.yuv",
- mDumpFrmCnt, dim.width, dim.height, frame->frame_idx);
+ snprintf(buf, sizeof(buf), "%dr_%dx%d_%d.raw",
+ mDumpFrmCnt,
+ dim.width,
+ dim.height,
+ frame->frame_idx);
}
break;
case QCAMERA_DUMP_FRM_JPEG:
{
mParameters.getStreamDimension(CAM_STREAM_TYPE_SNAPSHOT, dim);
- snprintf(buf, sizeof(buf), "/data/%dj_%dx%d_%d.yuv",
- mDumpFrmCnt, dim.width, dim.height, frame->frame_idx);
+ snprintf(buf, sizeof(buf), "%dj_%dx%d_%d.yuv",
+ mDumpFrmCnt,
+ dim.width,
+ dim.height,
+ frame->frame_idx);
}
break;
default:

 root@msm8226:/data # ls |grep isp
ls |grep isp
isp_dump_64_preview_640_480.yuv
isp_dump_65_preview_640_480.yuv
isp_dump_66_preview_640_480.yuv
isp_dump_67_preview_640_480.yuv
isp_dump_68_preview_640_480.yuv
isp_dump_69_preview_640_480.yuv
isp_dump_70_preview_640_480.yuv
isp_dump_71_preview_640_480.yuv
isp_dump_72_preview_640_480.yuv
isp_dump_73_preview_640_480.yuv
isp_dump_74_preview_640_480.yuv

[13]

<3>[  653.808032] mmc1: data txfr (0x00200000) error: -84 after 16 ms
<6>[  653.808039] sdhci: =========== REGISTER DUMP (mmc1)===========
<6>[  653.808049] sdhci: Sys addr: 0x00000000 | Version:  0x00003a02
<6>[  653.808060] sdhci: Blk size: 0x00007200 | Blk cnt:  0x00000001
<6>[  653.808070] sdhci: Argument: 0x00504590 | Trn mode: 0x00000003
<6>[  653.808080] sdhci: Present:  0x03e80100 | Host ctl: 0x00000017
<6>[  653.808090] sdhci: Power:    0x0000000f | Blk gap:  0x00000000
<6>[  653.808101] sdhci: Wake-up:  0x00000000 | Clock:    0x00000007
<6>[  653.808111] sdhci: Timeout:  0x0000000f | Int stat: 0x00000000
<6>[  653.808122] sdhci: Int enab: 0x03ff800b | Sig enab: 0x03ff800b
<6>[  653.808130] sdhci: AC12 err: 0x00000000 | Slot int: 0x00000000
<6>[  653.808141] sdhci: Caps:     0x2329c8b2 | Caps_1:   0x00008007
<6>[  653.808151] sdhci: Cmd:      0x0000183a | Max curr: 0x00000000
<6>[  653.808161] sdhci: Resp 1:   0x00000000 | Resp 0:   0x00000900
<6>[  653.808171] sdhci: Resp 3:   0x00000000 | Resp 2:   0x00000000
<6>[  653.808178] sdhci: Host ctl2: 0x0000000b
<6>[  653.808187] sdhci: ADMA Err: 0x00000000 | ADMA Ptr: 0x33b60010
<6>[  653.808196] mmc1: clk: 200000000 clk-gated: 0 claimer: mmcqd/1 pwr: 14
<6>[  653.808203] mmc1: rpmstatus[pltfm](runtime-suspend:usage_count:disable_depth)(0:0:0)
<6>[  653.808209] sdhci: ===========================================

 

Patch for T card tx error.
   
    Change-Id: I4be8a30fbee0b6c06666f6d48d70a622f01d7d82

diff --git a/drivers/mmc/card/block.c b/drivers/mmc/card/block.c
index 743668b..e36a2cf 100755
--- a/drivers/mmc/card/block.c
+++ b/drivers/mmc/card/block.c
@@ -2326,8 +2326,11 @@ static int mmc_blk_cmd_err(struct mmc_blk_data *md, struct mmc_card *card,
   u32 blocks;
   if (!brq->data.fault_injected) {
    blocks = mmc_sd_num_wr_blocks(card);
-   if (blocks != (u32)-1)
+   if (blocks != (u32)-1) {
+    if (blocks << 9 == blk_rq_bytes(req))
+     pr_err("%s: buggy SD card\n", __func__);
     ret = blk_end_request(req, 0, blocks << 9);
+   }
   } else
    ret = blk_end_request(req, 0, brq->data.bytes_xfered);
  } else {
@@ -2488,29 +2491,61 @@ static int mmc_blk_issue_rw_rq(struct mmc_queue *mq, struct request *rqc)
     rqc = NULL;
     goto cmd_abort;
    }
+   if (ret)
+    pr_warning("%s: %s: command partially completed\n",
+     mmc_hostname(card->host), __func__);
    break;
   case MMC_BLK_CMD_ERR:
    ret = mmc_blk_cmd_err(md, card, brq, req, ret);
-   if (!mmc_blk_reset(md, card->host, type))
+   if (!mmc_blk_reset(md, card->host, type)) {
+    if (!ret) {
+     BUG_ON(card->host->areq);
+     pr_err("%s: invalid scenario for MMC_BLK_CMD_ERR, xfered %d, req %p\n",
+       __func__,
+       brq->data.bytes_xfered,
+       req);
+     goto start_new_req;
+    }
+    pr_err("%s: %s: reset success, retry prev command\n",
+     mmc_hostname(card->host), __func__);
     break;
+   }
+   pr_err("%s: %s: reset failed, aborting command\n",
+     mmc_hostname(card->host), __func__);
    goto cmd_abort;
   case MMC_BLK_RETRY:
-   if (retry++ < 5)
+   if (retry++ < 5) {
+    pr_err("%s: %s: MMC_BLK_RETRY, cnt %d, ret %d\n",
+     mmc_hostname(card->host),
+     __func__, retry, ret);
     break;
+   }
    /* Fall through */
   case MMC_BLK_ABORT:
-   if (!mmc_blk_reset(md, card->host, type))
+   if (!mmc_blk_reset(md, card->host, type)) {
+    pr_err("%s: %s: reset success during abort, ret %d\n",
+     mmc_hostname(card->host),
+     __func__, ret);
     break;
+   }
+   pr_err("%s: %s: reset failed during abort, aborting command\n",
+     mmc_hostname(card->host), __func__);
    goto cmd_abort;
   case MMC_BLK_DATA_ERR: {
    int err;
 
    err = mmc_blk_reset(md, card->host, type);
-   if (!err)
+   if (!err) {
+    pr_err("%s: %s: reset success, during data err ret %d\n",
+     mmc_hostname(card->host),
+     __func__, ret);
     break;
+   }
    if (err == -ENODEV ||
     mq_rq->packed_cmd != MMC_PACKED_NONE)
     goto cmd_abort;
+   pr_err("%s: %s: reset failed, during data err fall through\n",
+     mmc_hostname(card->host), __func__);
    /* Fall through */
   }
   case MMC_BLK_ECC_ERR:
@@ -2528,10 +2563,16 @@ static int mmc_blk_issue_rw_rq(struct mmc_queue *mq, struct request *rqc)
     */
    ret = blk_end_request(req, -EIO,
       brq->data.blksz);
+   pr_err("%s: %s: reset failed for ecc err, ret %d\n",
+     mmc_hostname(card->host),
+     __func__, ret);
+
    if (!ret)
     goto start_new_req;
    break;
   case MMC_BLK_NOMEDIUM:
+   pr_err("%s: %s: no medium, aborting command\n",
+     mmc_hostname(card->host), __func__);
    goto cmd_abort;
   default:
    pr_err("%s:%s: Unhandled return value (%d)",

 

diff --git a/drivers/mmc/card/queue.c b/drivers/mmc/card/queue.c
index 8986829..637c467 100755
--- a/drivers/mmc/card/queue.c
+++ b/drivers/mmc/card/queue.c
@@ -30,6 +30,9 @@
  */
 #define DEFAULT_NUM_REQS_TO_START_PACK 17
 
+struct mmc_queue *tsr_mq;
+struct mmc_card *tsr_card;
+
 /*
  * Prepare a MMC request. This just filters out odd stuff.
  */
@@ -61,6 +64,11 @@ static int mmc_queue_thread(void *d)
 
  current->flags |= PF_MEMALLOC;
 
+ if (mmc_card_sd(card)) {
+  tsr_mq = mq;
+  tsr_card = card;
+ }
+
  down(&mq->thread_sem);
  do {
   struct mmc_queue_req *tmp;
@@ -94,6 +102,18 @@ static int mmc_queue_thread(void *d)
     * Current request becomes previous request
     * and vice versa.
     */
+   if (mmc_card_sd(card) && mq->mqrq_prev->req) {
+    int in_flight = 0;
+    in_flight = q->in_flight[0] +
+     q->in_flight[1];
+    if (in_flight >= 2) {
+     pr_err("TSR %s: in_flight %d\n", __func__, in_flight);
+     BUG_ON(1);
+    } else if (in_flight == 1 && !mq->mqrq_cur->req) {
+     pr_err("TSR %s: in_flight %d\n", __func__, in_flight);
+     BUG_ON(1);
+    }
+   }
    mq->mqrq_prev->brq.mrq.data = NULL;
    mq->mqrq_prev->req = NULL;
    tmp = mq->mqrq_prev;

 

[15]

 [16]

屏无法唤醒

 

1、DDR部分原理图和LAYOUT检查，确认没有问题。
2、CPU和memory部分电压和时钟做了测量，没有看到有异常
3、换了一颗hynix的memory，故障现象没有消失
4、换了一颗CPU，刚开始开机正常，跑了10几分钟DDR测试工具死机。再开机再跑几十秒就死了，再后来开不了机了。重新对CPU加焊了下，目前开关机10几次正常，正准备再次跑DDR测试。

 

[17]


 

 

 


 [18]尝试修改内存信号---调试

diff --git a/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c b/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c
index 7e9005f..105c957 100644
--- a/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c
+++ b/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c
@@ -172,7 +172,7 @@ uint32 ddr_dqphy_config_MSM8x10[][2] =
   {HWIO_ADDR(PHY_DQ_ADDR(RD_CDC_SW_OVRD_CFG)), 0x00000},
   {HWIO_ADDR(PHY_DQ_ADDR(TOP_CFG)), 0x00000031},
   {HWIO_ADDR(PHY_DQ_ADDR(WR_CDC_DELAY_CFG)), 0x0000758},
-  {HWIO_ADDR(PHY_DQ_ADDR(WR_CDC_OFFSET_CFG)), 0x000000},
+  {HWIO_ADDR(PHY_DQ_ADDR(WR_CDC_OFFSET_CFG)), 0x000086},
   {HWIO_ADDR(PHY_DQ_ADDR(WR_CDC_SW_OVRD_CFG)), 0x00000},
   {0x0, 0x0}
 };

    [patch]add DDR fault tolerant ability

diff --git a/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c b/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c
index cb95d40..7e9005f 100644
--- a/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c
+++ b/boot_images/core/boot/ddr/hw/msm8x10/ddr_config.c
@@ -143,8 +143,8 @@ uint32 ddr_caphy_config_MSM8x10[][2] =
   {HWIO_ADDR(PHY_CA_ADDR(IOC_CTLR_CHAR_CFG)), 0x00000000},
   {HWIO_ADDR(PHY_CA_ADDR(IOC_CTLR_PNCNT_CFG)), 0x00001010},
   {HWIO_ADDR(PHY_CA_ADDR(IOC_CTLR_TIMER_CFG)), 0xFA000020},
-  {HWIO_ADDR(PHY_CA_ADDR(PAD_CFG0)), 0x20222240},
-  {HWIO_ADDR(PHY_CA_ADDR(PAD_CFG1)), 0x20220440},
+  {HWIO_ADDR(PHY_CA_ADDR(PAD_CFG0)), 0x202222C0},
+  {HWIO_ADDR(PHY_CA_ADDR(PAD_CFG1)), 0x202204C0},
   {HWIO_ADDR(PHY_CA_ADDR(PAD_CFG2)), 0x10000000},
   {HWIO_ADDR(PHY_CA_ADDR(PAD_CFG3)), 0x313003FF},
   {HWIO_ADDR(PHY_CA_ADDR(PAD_CFG4)), 0x303003FF},
@@ -161,11 +161,11 @@ uint32 ddr_dqphy_config_MSM8x10[][2] =
   {HWIO_ADDR(PHY_DQ_ADDR(DQS_IOC_SLV_CFG)), 0x80000000},
   {HWIO_ADDR(PHY_DQ_ADDR(DQ_IOC_SLV_CFG)), 0x80000000},
   {HWIO_ADDR(PHY_DQ_ADDR(DQ_SMT_STATUS)), 0x00000000},
-  {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG0)), 0xA0222240},
-  {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG1)), 0xA0222240},
+  {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG0)), 0xA02222C0},
+  {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG1)), 0xA02222C0},
   {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG2)), 0x10000000},
   {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG3)), 0x1000FF11},
-  {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG4)), 0x20222240},
+  {HWIO_ADDR(PHY_DQ_ADDR(PAD_CFG4)), 0x202222C0},
   {HWIO_ADDR(PHY_DQ_ADDR(RD_CDC_DELAY_CFG)), 0x00003AC},
   {HWIO_ADDR(PHY_DQ_ADDR(RD_CDC_OFFSET_CFG)), 0x000000},
   {HWIO_ADDR(PHY_DQ_ADDR(RD_CDC_SLAVE_DDA_CFG)), 0x16334},

 

 

 

SDM439平台出现部分机型SD卡不能识别mmc1: error -110 whilst initialising SD card【学习笔记】...
05-27 1983
SDM439平台出现部分机型SD卡不能识别mmc1: error -110 whilst initialising SD card 打印了如下的log： 1 05-10 18:01:02.699 <3>[ 3079.000640] mmc1: data txfr (0x00100000) error: -110 after 163 ms 2 05-10 18:0...
高通Camera 驱动调试要点（二）
热门推荐
ForDream321的专栏
10-26 1万+
这篇文章主要介绍数据流这边Camera所遇到的问题，主要介绍bus overflow和sof freeze； （一）bus overflow   摄像头传感器时钟通道(即 MIPI DDR 时钟)的工作频率与激活的数据通道数决定摄像头传感器在指定操作模式下的总数据传输速率(吞吐量)。每个通道的数据传输速率是 MIPI DDR 时钟速度的两倍。例如,工作在 200 MHz MIPI DD
Arm内核的Oops错误定位方法
boyemachao的专栏
06-29 7569
出错的log信息如下: 1 Unable to handle kernel NULL pointer dereference at virtual address 00000014 2 pgd = c0004000 3 [00000014] *pgd=00000000 4 Internal error: Oops: 5 [#1] PREEMPT SMP ARM 5 Modules linked in: 6 CPU: 0 PID: 1 Comm: swapper/0 Not tainted 3
高通平台摄像头调试技巧
jinron10的专栏
12-10 4866
1、Camera启动速度(Hal层的耗时) Camera的启动速度，是我们经常做优化的地方，除了通过仪器去测量，还可以通过Log来查看！搜索关键字：[KPI Perf] 01-24 17:55:57.657 371 4074 D QCamera2HWI: [KPI Perf] int qcamera::QCamera2HardwareInterface::openCamera(hw_d...
主板死机引发对complete的学习
Android开发
09-19 2781
static irqreturn_t send_handle(int irq, void *dev_id) { complete(&tx_done_complete); } int xxx_probe(){ ret = request_irq(virq,send_handle,IRQF_TRIGGER_RISING,"cmt2300a_send_complete", NULL); init_completion(&tx_done_complete); }
qualcomm camera 学习笔记
入门虽易，精通不易，且学且珍惜！
06-16 1万+
高通camera学习笔记概述（一二三四五） http://www.cnblogs.com/potato-coffee/p/4172009.html android camera 预览 app到hal  http://blog.csdn.net/tankai19880619/article/details/22795217 电路图，时序图，移植，介绍参数 移植时用  http://www.cnb
Qualcomm MSM8937 dual DSI 笔记
素炒瓢儿白的CSDN博客
02-14 2498
简述 在 MSM8937 上有支持了 dual dsi 功能，本人在 msm8937-android6.0 上将其调通，以下简记其实现的思路以及关键代码片段。 高通支持的双屏有两种方式：其一为将一副图片左右均分，然后通过两个 DSI 硬件接口刷到屏幕上去，似乎在 MSM8952 上就是这种方式；其二为将一副图片不做分割操作，直接通过两个 DSI 硬件接口刷到屏幕上去，在 MSM8937 上则是支持...
deepin驱动精灵_解决宏基笔记本没有Qualcomm_Atheros_QCA9377无线驱动
weixin_42291186的博客
01-13 4068
本帖最后由 Iczelion 于 2016-7-9 22:12 编辑硬件：宏基笔记本Acer K50-10-53TR系统：Deepin GNU/Linux 2015版今天帮同学在他的笔记本上安装了Win10专业版和Deepin GNU/Linux 2015版双系统，安装后选择进入Deepin GNU/Linux发现只有有线链接，没有无线网卡驱动。以下是在网上找到的解决方案。1.首先进入Win10，...
Qualcomm Atheros AR9462 WiFi Adapter_10.0.3.456(含Killer N1202、AR5B22).zip
01-22
标题中的"Qualcomm Atheros AR9462 WiFi Adapter_10.0.3.456(含Killer N1202、AR5B22).zip"指的是一个针对Qualcomm Atheros AR9462无线网卡的驱动程序更新包。此驱动程序版本为10.0.3.456，并且包含了对Killer N1202...
Qualcomm-Atheros-QCA9377-main.zip
09-07
【Qualcomm Atheros QCA9377 主要特性与技术详解】 Qualcomm Atheros QCA9377是一款由高通（Qualcomm）子公司Atheros Communications开发的无线网络芯片，广泛应用于笔记本电脑、路由器和其他无线设备中。该芯片以...
驱动调试之段错误分析_根据oops信息确定出错的代码位置
u012830148的博客
03-26 3060
在内核调试过程中，经常会出现kernel painc等错致命错误导致内核崩溃，遇到这种问题可以根据oops信息逐步推断出内核出错的位置，找出出错的代码。 驱动程序一般分为作为模块通过 insmod加载 和直接编译进内核中两种。 利用tiny4412开发板，故意在驱动程序中引入一个指针错误。不进行ioremap，直接访问物理地址，这是非法的。 第一种：通过insmod加载 shell@tin...
根据内核Oops 定位代码
dolphin98629的专栏
04-03 1016
内核开发时有时候出现Oops，例如一个野指针会导致内核崩溃，如运行时出现以下log：现在有三种方法可以找出具体出现野指针的地方    5.438972] bells bells:  wm5102-aif1  samsung-i2s.0 mapping ok   [    5.443812] bells bells: Failed to add route OPCLK->Su
Android工程之SD卡问题
Tree_in_sea的博客
01-11 1884
/*****************************************************************************  * Author : Elvins Fu    yeyecheng_93@163.com  *  * Info : xxx Inc,(C) 2018-01-11, All rights revseved.  *  * Descri
高通（Qualcomm）LK源码深度分析（三）
sdkdlwk的博客
10-20 2982
本编文章的内容主要是分析 boot/recovery 的启动过程，其中的 boot 就是 android 的kernel, 是整个 android 系统的核心。本文的分析是紧接着 aboot_init 的分析内容的，只是因为其重要性才单独成为一章。 以下是前两篇文章回顾： 高通（Qualcomm）LK源码深度分析 高通（Qualcomm）LK源码深度分析（二） recovery boo
高通camera mipi不通的解决方法
ramiel_yang的专栏
03-30 5990
打开mipi中断的状态开关，也就是一个注册mipi中断的软件上的宏定义。 首先你要确定你的camera是几lane的，如果lane是对的，检查下模组，比如是2M的camera，是否真的使用了2lane，模组内部可能会断路使用一个lane。 其次你的平台配置如何，mipi的配置是否在stream on之前，中断状态是否有？ 再次，mipi的配置高通7x和8x是不一样的，走csi和csid的流程
linux kernel note
☆ 默默地牛逼着
08-02 3140
1.?????v1 #define CHARGER_PRINT(fmt,arg...) printk(" "fmt"\n",##arg)v2 #define CHARGER_PRINT(fmt,arg...) printk(" "fmt"\n
高通apq8074修改DDR频率
ruidongren的专栏
10-12 3216
1. M odify SBL1 first. The calculation is 19.2*(L+(M/N))/2 //时钟计算方法 /*-----------------------------------------------------------------------*/ /* GPLL3 - BIMC PLL */ /*----------------
Linux: 运行存在的可执行文件提示 No such file or directory
JiMoKuangXiangQu的专栏
11-10 2万+
linux, 文件系统, ftrace
camera log分析
齐海谦的专栏
09-28 1万+
一、关于Camera 方向 跟 LCD 方向的配置问题: 物理硬件上: 因为 sensor 是矩形(长方形)，分长和宽,长宽比通常是 4:3，且 sensor 扫描方向都是按长边扫描，以2M芯片为例，芯片从 (1,1) 像素点开始曝光，依次到终点(1200,1600). 而 LCD 有两种扫描方式: 沿短边扫描（物理竖屏） 和 沿长边扫描（物理横屏）。 沿短边: (1,1) -
Android开发入门指南与学习笔记
此外，半导体公司如Intel、Qualcomm、Samsung、Nvidia等也在联盟中扮演重要角色，他们为Android设备提供了先进的处理器和芯片解决方案，推动了Android设备的性能提升和技术创新。 在Android开发过程中，开发者需要...
评论
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值