bitbake -v meta-toolchain
2020年2月27日 星期四
2020年1月16日 星期四
Signal and Encoding of USB System (Part 5/6)
Source https://www.engineersgarage.com/tutorials/signal-and-encoding-of-usb-system-part-5-6/
In this series, USB requests and descriptors have been already discussed. With the knowledge of USB descriptors, requests, transfer types, and transactions it is now easy to understand how data is communicated over USB interface. Another aspect of the USB communication is line states and encoding, which is generally ignored by the system designers as a USB device can be designed without knowing the Signaling and Encoding schemes. The signaling and encoding schemes are implemented at the lower levels (usually on the controller IC) and usually gets abstracted during system designing. However, understanding the terms related to bus states and encoding would help in getting a comprehensive view of the USB System.
Bus States
The Bus state refers to the signal voltage level on the USB bus or the conditions that these voltages signify -:
Differential 0 and Differential 1: The Differential 0 state refers to the condition when D+ data line is logic low and D- is logic high. The Differential 1 state occurs when D+ is logic high and D- is logic low.
Single Ended Zero: This state is also known as SE0. It occurs when both D+ and D- are logic low. The SE0 state is used by bus when entering the Disconnect, Reset and EOP (End of Packet) states.
Single Ended One: Also known as SE1. It occurs when both D+ and D- are logic high. This state can be taken as a compliment of the SE0 state. SE1 should never occur as it is an invalid bus State.
Idle: This state occurs when bus remains idle or there is no exchange of packet. For a full speed device, the voltage at D+ will be more positive than D- while for a low-speed device, the voltage at D- will be more positive than D+
Data J and Data K: Data J and K states are used to define the Data States like the Differential 0 and 1 that refers to the bus states. They are defined by Differential 0 or 1 state and the speed of the bus.
Fig. 1: Table listing Flip-Flop Status for Differential 0 and Differential 1 Bus States
The data states are used to describe an event even when the voltages at data lines are differential. For example, a Start of Packet state exists when the bus changes from Idle to the K state.
Start of Packet: Also known as SOP. This state occurs when there is a change from Idle to K state. Every transmission of the packet begins with SOP.
End of Packet: Also known as EOP. This state occurs when SEO state occurs for two-bit times, followed by a J state for 1-bit time.
Disconnect: A downstream port at which the device is connected enters disconnect state when an SE0 state has lasted for at least 2.5 uS.
Connect: A downstream port enters connect state when there is an Idle state for a minimum 2.5 uS and not more than 2 mS
Some Other States and Signals
Keep Alive Signal: This signal is represented by a low-speed EOP. The main motive of this signal is to keep the device in an idle state. This signal is sent at least once every millisecond, to keep the device from entering suspend state.
Suspend State: The device enters Suspend state when there is no exchange of packet or the bus remains idle for 3 mS. This state has been incorporated in USB for power conservation. In Suspended state, the device must not draw more than 500 uA of current. A suspended device must recognize the reset and resume signal.
Resume: Resume signal is used to wake the device from Suspend state. The Host wakes the device from suspend state by keeping the bus in K state for at least 20 ms followed by a low-speed EOP.
Reset State: A device enters Reset state if SE0 is applied for more than 2.5 uS. The reset sets the device to its default unconfigured state.
Detached State: This state occurs when the USB device is detached from the Host. In this state, the Host sees both data lines low.
Attached State: This state occurs when the USB device is attached to the Host. The host recognizes this state if it sees either D+ or D- high.
Fig. 2: Table listing conditions for USB Bus States
Fig. 3: Signal Diagram for different USB Bus States
Speed Identification
In USB System, the operating speed of the device is detected by checking which data line is pulled up. The data line is pulled up to 3.3V derived from voltage line. Normally, 1.5-kilo ohm resistor is used for a pull-up.
For Low speed, D- data line is pulled up. While for full speed, D+ data line is pulled up. High-speed devices are initially detected as full speed devices.
Encoding Scheme
The USB employs NRZI (Non Return to Zero Inversion) encoding mechanism to encode the data on the bus. In NRZI encoding, a ‘1’ is represented by no change in level while a ‘0’ is represented by change in level. Together with NRZI encoding, bit stiffing and SYNC field is used for synchronization between host and device.
Fig. 4: Signal Diagram illustrating Non Return to Zero Inversion (NRZI) Encoding Scheme
Bit stiffing
When long series of zeros are transmitted using NRZI, it causes a transition in the levels. But when long series of one’s is transmitted, no transition takes place as per NRZI encoding scheme. No transition in levels for a long time can confuse the receiver and makes it desynchronized.
Bit stiffing is a process in which a zero is inserted in raw data after every six consecutive ones. The inserting of zero causes transition in level. The receiver must recognize the stuffed bits and discard them after decoding the NRZI data.
In case if no transition takes place in NRZI signal after six consecutive one’s, then the receiver decides that bit stiffing has not been done and discards the data received.
Fig. 5: Signal Diagram showing Bit Shifting in Non Return to Zero Inversion (NRZI) Encoding Scheme
SYNC field
The bit stiffing alone is not enough to take care of the synchronization between transmitter and receiver. To keep transmitter and receiver synchronized, SYNC field is used. In this mechanism, each packet begins with a SYNC field which enables the receiver to synchronize the clock
For low/full speed, SYNC field is of 8 bits: KJKJKJKK. The first k serves as Start of Packet. For high speed, SYNC field is of 32 bits: fifteen KJ repetitions, followed by KK. The alternate Ks and Js provide transitions for synchronizing and the final two KKs mark the end of SYNC field.
End of Packet
The End of Packet that is SE0 for 2-bit times followed by J state for 1-bit time marks the end of packet.
This is all needed to know about the signal and encoding schemes implemented in the USB system. In the final part of the series, selection of a controller chip for developing USB enabled custom devices is discussed.
2020年1月3日 星期五
Andriod version, API level, and Linux version
Source https://android.stackexchange.com/questions/51651/which-android-runs-which-linux-kernel
Linux version in AOSP
Android Version |API Level |Linux Version in AOSP |Header Version
----------------------------------------------------------------------------------------
1.5 Cupcake |3 |(2.6.27) |
1.6 Donut |4 |(2.6.29) |2.6.18
2.0/1 Eclair |5-7 |(2.6.29) |2.6.18
2.2.x Froyo |8 |(2.6.32) |2.6.18
2.3.x Gingerbread |9, 10 |(2.6.35) |2.6.18
3.x.x Honeycomb |11-13 |(2.6.36) |2.6.18
4.0.x Ice Cream San|14, 15 |(3.0.1) |2.6.18
4.1.x Jelly Bean |16 |(3.0.31) |2.6.18
4.2.x Jelly Bean |17 |(3.4.0) |2.6.18
4.3 Jelly Bean |18 |(3.4.39) |2.6.18
4.4 Kit Kat |19, 20 |(3.10) |2.6.18
5.x Lollipop |21, 22 |(3.16.1) |3.14.0
6.0 Marshmallow |23 |(3.18.10) |3.18.10
7.0 Nougat |24 | 3.18.48 4.4.0 |4.4.1
7.1 Nougat |25 | ? |4.4.1
8.0 Oreo |26 | 3.18.72 4.4.83 4.9.44 |4.10.0
8.1 Oreo |27 | 3.18.70 4.4.88 4.9.56 |4.10.0
9.0 Pie |28 | 4.4.146 4.9.118 4.14.61 |4.15.0
10.0 Q |29 | 4.9.191 4.14.142 4.19.71|5.0.3
Table compiled from:
- Wikipedia Android version history (in parentheses; can't find its source)
- android / kernel / common / Makefile (the Linux in AOSP)
- android / platform / external / kernel-headers / original / uapi / linux / version.h (the Linux headers as used by Bionic in AOSP)
Note: other “distributions” of Android might run different kernel versions.
Anyone is welcome to correct and expand this table. Please keep formatting.
Why Android OS/kernel version mismatches
OEMs tend to ship their major Android updates with a similar kernel to the one the device was initially released with. This is most likely to avoid hardware and driver incompatibility issues with a newer kernel. For example, a device that has been updated to run Android 8.0 Oreo but originally shipped with 6.0 Marshmallow may still be running Linux 3.18.x instead of Linux 4.10.x.
Vendor Test Suite (VTS) requirements on Linux version
According to a 2017 article on XDA-Developers (shortened):
One aspect that has been up to the OEM is the Linux kernel version but this is changing with Android Oreo.As long as the OEM was able to pass the certification tests that Google lays out, then they didn’t care what kernel version was used in a new device. This generally wasn’t an issue as most OEMs would use the same version of the kernel for that generation that other OEMs were using, as it is tied heavily to what the hardware drivers support. However, some had been falling through the cracks and this started to cause security issues. This is something that Google has been taking seriously lately so it makes sense that they would want to start mandating this.Starting this year with smartphones which ship with Android Oreo, Google is requiring that all SoCs productized in 2017 must launch with kernel 4.4 or newer.Currently existing devices that are upgraded to Android Oreo are only required to run kernel version 3.18 or newer, and will not have to be upgraded to support Project Treble.
According to https://android.googlesource.com/platform/test/vts-testcase/kernel/, there was a requirement on the Linux version, VtsKernelConfigTest or VtsKernelVersionTest, on 8.0, 8.1, and 9.0, but later removed on 10.0. For 8.0 and 8.1, it requires the version to be either 3.18, 4.4, or 4.9. For 9.0, it requires the version to be at least 4.4.107, 4.9.84, or 4.14.42.
More technical and recent information available directly from the AOSP website: https://source.android.com/devices/architecture/kernel/core-kernel-reqs
2019年12月29日 星期日
Manage file permissions on Unix-like systems
Source https://kb.iu.edu/d/abdb
On this page:
- Overview
- View file permissions
- Change file permissions
- Common issues when sharing data with other users
- Get help
Overview
Unix-like operating systems, such as Linux, running on shared high-performance computers use settings called permissions to determine who can access and modify the files and directories stored in their file systems. Each file and directory in a file system is assigned "owner" and "group" attributes.
Most commonly, by default, the user who creates a file or directory is set as owner of that file or directory. When needed (for example, when a member of your research team leaves), the system's root administrator can change the user attribute for files and directories.
The group designation can be used to grant teammates and/or collaborators shared access to an owner's files and directories, and provides a convenient way to grant access to multiple users.
View file permissions
To view the permissions for all files in a directory, use the
ls command with the -la options. Add other options as desired; for help, see List the files in a directory in Unix.
For example, if you enter:
ls -lah
You should see output similar to the following:
-rw-r--r-- 1 user1 group1 62 Jan 15 16:10 myfile.txt drwxr-xr-x 2 user1 group1 2048 Jan 15 17:10 Example
In the output example above, the first character in each line indicates whether the listed object is a file or a directory. Directories are indicated by a (
d); the absence of a d at the beginning of the first line indicates that myfile.txt is a regular file.
The letters
rwx represent different permission levels:| Permission | Files | Directories |
|---|---|---|
r | can read the file | can ls the directory |
w | can write the file |
can modify the directory's contents
|
x | can execute the file | can cd to the directory |
Note the multiple instances of
r, w, and x. These are grouped into three sets that represent different levels of ownership:- Owner or user permissions: After the directory (
d) slot, the first set of three characters indicate permission settings for the owner (also known as the user).In the example-rw-r--r--, the owner permissions arerw-, indicating that the owner can read and write to the file but can't execute it as a program.In the exampledrwxr-xr-x, the owner permissions arerwx, indicating that the owner can view, modify, and enter the directory. - Group permissions: The second
rwxset indicates the group permissions. In the fourth column of the example above,group1is the group name.In the example-rw-r--r--, group members can only read the file.In the exampledrwxr-xr-x, group members can view as well as enter the directory. - Other permissions: The final
rwxset is for "other" (sometimes referred to as "world"). This is anyone outside the group. In both examples above, these are set to the same permissions as the group.
Change file permissions
To change file and directory permissions, use the command
chmod (change mode). The owner of a file can change the permissions for user (u), group (g), or others (o) by adding (+) or subtracting (-) the read, write, and execute permissions.
There are two basic ways of using
chmod to change file permissions: The symbolic method and the absolute form.Symbolic method
The first and probably easiest way is the relative (or symbolic) method, which lets you specify permissions with single letter abbreviations. A
chmod command using this method consists of at least three parts from the following lists:| Access class | Operator | Access Type |
|---|---|---|
| u (user) | + (add access) | r (read) |
| g (group) | - (remove access) | w (write) |
| o (other) | = (set exact access) | x (execute) |
| a (all: u, g, and o) |
For example, to add permission for everyone to read a file in the current directory named
myfile, at the Unix prompt, enter:chmod a+r myfile
The
a stands for "all", the + for "add", and the r for "read".
Note:
This assumes that everyone already has access to the directory where
myfile is located and its parent directories; that is, you must set the directory permissions separately.
If you omit the access class, it's assumed to be all, so you could also enter the previous example as:
chmod +r myfile
You can also specify multiple classes and types with a single command. For example, to remove read and write permission for group and other users (leaving only yourself with read and write permission) on a file named
myfile, you would enter:chmod go-rw myfile
You can also specify that different permissions be added and removed in the same command. For example, to remove write permission and add execute for all users on
myfile, you would enter:chmod a-w+x myfile
In each of these examples, the access types that aren't specified are unchanged. The previous command, for example, doesn't change any existing settings specifying whether users besides yourself may have read (
r) access to myfile. You could also use the exact form to explicitly state that group and other users' access is set only to read with the = operator:chmod go=r myfile
The
chmod command also operates on directories. For example, to remove write permission for other users on a subdirectory named mydir, you would enter:chmod o-w mydir
To do the same for the current directory, you would enter:
chmod o-w
To change permissions recursively in all subdirectories below the specified directory, add the
-R option; for example, to grant execution permissions for other users to a directory (mydir) and all the subdirectories it contains, you would enter:chmod -R o+x mydir
Be careful when setting the permissions of directories, particularly your home directory; you don't want to lock yourself out by removing your own access. Also, you must have execute permission on a directory to switch (
cd) to it.Absolute form
The other way to use the
chmod command is the absolute form, in which you specify a set of three numbers that together determine all the access classes and types. Rather than being able to change only particular attributes, you must specify the entire state of the file's permissions.
The three numbers are specified in the order: user (or owner), group, and other. Each number is the sum of values that specify read, write, and execute access:
| Permission | Number |
|---|---|
| Read (r) | 4 |
| Write (w) | 2 |
| Execute (x) | 1 |
Add the numbers of the permissions you want to give; for example:
- For file
myfile, to grant read, write, and execute permissions to yourself (4+2+1=7), read and execute permissions to users in your group (4+0+1=5), and only execute permission to others (0+0+1=1), you would use:chmod 751 myfile
- To grant read, write, and execute permissions on the current directory to yourself only, you would use:
chmod 700
You can think of the three digit sequence as the sum of attributes you select from the following table:
| Read by owner | 400 |
| Write by owner | 200 |
| Execute by owner | 100 |
| Read by group | 040 |
| Write by group | 020 |
| Execute by group | 010 |
| Read by others | 004 |
| Write by others | 002 |
| Execute by others | 001 |
Sum all the accesses you wish to permit. For example, to give write and execute privileges to the owner of
myfile (200+100=300), and give read privileges to all (400+040+004=444), you would enter:chmod 744 myfile
Some other examples are:
| 777 | anyone can do anything (read, write, or execute) |
| 755 | you can do anything; others can only read and execute |
| 711 | you can do anything; others can only execute |
| 644 | you can read and write; others can only read |
Common issues when sharing data with other users
Important:
Be sure you understand your responsibilities when processing, storing, and sharing data containing protected health information (PHI). For more, see Your legal responsibilities for protecting data containing protected health information (PHI) when using UITS Research Technologies systems and services.
To share a file or directory that you own with someone, you can grant read and execute privileges for that user. However, you must also set the same privileges on any parent directories above the item you're sharing; if you don't, the user can't look and change into (
cd) all the parent directories above your file or directory.
If you think of a file system as a physical place, then permissions work like keys that let you access different directories:
- The read (
r) permission lets users look (ls) into directories. - The execute (
x) permission lets users move (cd) into directories. - The write (
w) permission lets users add and remove files.
For example, say you want to give someone access to
/N/u/username/Carbonate/scripts. Imagine the path as a physical space:/Nis the gated community where you live./uis the unit./usernameis your apartment./Carbonateis a room in your apartment./scriptsis a closet in your room.
If someone wanted to run your scripts, you would need to give that person access to every part of
/N/u/username/Carbonate/scripts. You might try to do it this way:chmod +rx /N/u/username/Carbonate/scripts
However, a user can't read or access a subdirectory unless the user also has
x permissions to the parent directories. In other words, the above command gives out a key to your closet, but not to your room or apartment.
To resolve this, give
x permissions to the parent directories you control:chmod +x /N/u/username/ chmod +x /N/u/username/Carbonate
This will let others move (
cd) to the scripts directory. Because the parent directories don't have r permissions, users will only be able to look (ls) within the scripts directory, keeping the rest of your file system private.Get help
For more about
chmod, consult the manual page. At the Unix prompt, enter:man chmod
At Indiana University, for personal or departmental Linux or Unix systems support, see Get help for Linux or Unix at IU.
This is document abdb in the Knowledge Base.
Last modified on 2019-08-23 15:27:34.
Last modified on 2019-08-23 15:27:34.
Contact us
For help or to comment, email the UITS Support Center.
訂閱：
文章 (Atom)