Time stamps can be used to associate a time with an event. A time stamp can reflect the actual time or can reflect an elapsed time relative to an arbitrary time origin. In some situations, it is desired to compare time stamps from two different devices. For example, in debugging and testing a System on Chip (SoC) device, external test equipment can be used to monitor events with respect to the System on Chip. The System on Chip can generate time stamps and associate them in a log with various events that occur internal to the System on Chip. Likewise, the external test equipment can generate its own time stamps and associate them in a log with various events that it detects in monitoring the System on Chip. It is often desired to correlate the logs to determine what events were occurring internal to the System on Chip at the time a certain external event was detected by the external test equipment. If the System on Chip and the external test equipment are on the same time domain, the time stamps of the different devices would be relative to the same time origin. However, if different time domains are used, the time stamps generated by the System on Chip and the external test equipment would be relative to different time origins, and, therefore, not correlated. Ad hoc comparison methods can be used to correlate the time stamps by checking for patterns of external activity.
Overview
By way of introduction, the below embodiments relate to a system and method for time stamp synchronization. In one embodiment, a method for time stamp synchronization between first and second devices is provided that is performed in the second device. The method comprises: receiving a first time stamp of the first device, wherein the first time stamp was generated in response to a time stamp synchronization event common to the first and second devices; generating a second time stamp of the second device in response to the time stamp synchronization event, wherein the first and second time stamps are in different time domains; and correlating the first and second time stamps, wherein correlating the first and second time stamps provide a relationship between the time domains because the first and second time stamps were both generated with respect to the time stamp synchronization event common to the first and second devices.
In some embodiments, the time stamp synchronization event is generated by the first device and sent to the second device.
In some embodiments, the time stamp synchronization event is part of a header of a communication frame sent from the first device to the second device.
In some embodiments, the first time stamp is generated using a clock in the first device, wherein the second time stamp is generated using a clock in the second, device, wherein the method further comprises using two successive time stamp synchronization events to determine a frequency relationship between the clocks in the first and second devices.
In some embodiments, the method further comprises determining the frequency relationship between the clocks in the first and second devices over time to account for drift in at least one of the clocks in the first and second devices.
In some embodiments, the first device comprises a System on Chip (SoC) device, and wherein the second device comprises test equipment configured to monitor the System on Chip (SoC) device.
In some embodiments, the System on Chip device is part of a storage system.
In some embodiments, a memory of the storage system comprises a three-dimensional memory.
In another embodiment, a device is provided comprising an interface configured to communicate with a second device using a communication frame comprising a header and a payload; and a controller in communication with the interface. The controller is configured to: include a time synch indicator in the header of the communication frame; generate a time stamp; include the time stamp in the payload of the communication frame; and send the communication frame to the second device.
In some embodiments, the controller is further configured to: detect the time synch indicator; and generate a system event tracer (SET) event in response to detecting the time synch indicator; wherein the controller is configured to generate the time stamp in response to the system event tracer (SET) event.
In some embodiments, the controller is further configured to include a time synch code in the payload of the communication frame.
In some embodiments, the device comprises a System on Chip (SoC) device, and wherein the second device comprises test equipment configured to monitor the System on Chip (SoC) device.
In some embodiments, the System on Chip device is part of a storage system.
In some embodiments, a memory of the storage system comprises a three-dimensional memory.
In another embodiment, a device is provided comprising: a time stamp generator configured to operate in a first time domain; and means for synchronizing the first time domain with a second time domain of a second device based on time stamps generated by the device and the second device in response to a common time synchronization signal.
In some embodiments, the means for synchronizing comprises means for correlating the time stamps of the device and the second device to provide a relationship between the first and second time domains.
In some embodiments, the device further comprises means for generating the common time synchronization signal.
In some embodiments, the device comprises a System on Chip (SoC) device, and wherein the second device comprises test equipment configured to monitor the System on Chip (SoC) device.
In some embodiments, the System on Chip device is part of a storage system.
In some embodiments, a memory of the storage system comprises a three-dimensional memory.
In another embodiment, a host is provided comprising at least one interface through which to communicate with a first computing device comprising a first clock operating in a first time domain and a second computing device comprising a second clock operating in a second time domain; and a processor configured to: store a common time stamp generated by a third clock operating in a third time domain in each of first and second log entries; send the first log entry to the first device, wherein the first device is configured to store a first additional time stamp in the first log entry; and send the second log entry to the second device, wherein the second device is configured to store a second additional time stamp in the second log entry; wherein the common time stamp and the first and second additional time stamps provide an indication of skew between the first and second additional time stamps.
In some embodiments, the first, second, and third clocks are synchronized by a time server, but wherein the skew nevertheless exists.
In some embodiments, the first and second computing devices are further configured to send the first and second log entries, respectively, to a database external to the first and second computing devices.
In some embodiments, the host further comprises a storage system, wherein one of the first and second logs is a log from the storage system.
In some embodiments, the storage system comprises a three-dimensional memory.
In some embodiments, at least one of the logs comprises a kernel (kmsg) log and a logcat log.
In some embodiments, storing the common time stamp in each of the first and second log entries is performed as a simultaneous operation.
In another embodiment, a method for time stamp synchronization is provided. The method comprises: performing the following in a host, wherein the host is part of a system comprising a first device and a second device, wherein each of the host and first and second devices comprises a respective clock operating in a respective time domain, wherein at least one of the clocks is not synchronized with respect to another clock: providing a common time stamp to first and second log entries using a clock in the host; sending the first and second log entries from the host to the first and second devices, respectively, wherein the first and second devices provide their own time stamp to the log entries, whereby the first and second log entries have two time stamps: the common time stamp provided by the host and a time stamp provided by one of the computing devices; and determining an amount of non-synchronization between the clocks using the two time stamps provided in the log entries.
In some embodiments, the first and second devices are configured to send the first and second log entries, respectively, to a database external to the first and second devices.
In some embodiments, the host comprises a storage system, and wherein one of the first and second logs entries is from a log in the storage system.
In some embodiments, the storage system comprises a three-dimensional memory.
In some embodiments, at least one of the logs comprises a kernel (kmsg) log and a logcat log.
In some embodiments, providing the common time stamp is a simultaneous operation.
In another embodiment, a host is provided comprising: at least one interface through which to communicate with a first computing device comprising a first clock operating in a first time domain and a second computing device comprising a second clock operating in a second time domain; means for storing a common time stamp generated by a third clock operating in a third time domain in each of first and second log entries; means for sending the first log entry to the first device, wherein the first device is configured to store a first additional time stamp in the first log entry; and means for sending the second log entry to the second device, wherein the second device is configured to store a second additional time stamp in the second log entry; wherein the common time stamp and the first and second additional time stamps provide an indication of skew between the first and second additional time stamps.
In some embodiments, the first, second, and third clocks are synchronized by a time server, but wherein the skew nevertheless exists.
In some embodiments, the first and second computing devices are further configured to send the first and second log entries, respectively, to a database external to the first and second computing devices.
In some embodiments, the host further comprises a storage system, and wherein one of the first and second logs is a log from the storage system.
In some embodiments, the storage system comprises a three-dimensional memory.
In some embodiments, at least one of the logs comprises a kernel (kmsg) log and a logcat log.
In some embodiments, assigning the same time stamp generated by the third clock to each of the first and second log entries is performed as a simultaneous operation.
Other embodiments are possible, and each of the embodiments can be used alone or together in combination. Accordingly, various embodiments will now be described with reference to the attached drawings.
Turning now to the drawings,
As shown in
The time stamp generators 15, 25 can be implemented in any suitable way, and both time stamp generators 15, 25 can be implemented the same way, or each can be implemented in a different way. For example, in one embodiment, one or both of the time stamp generators 15, 25 can be implemented with a hardware timer/counter that receives a clock signal from a clock in the device 10, 20, which causes the timer/counter to count up with each clock pulse. The generated time stamps can be stored in an internal register or memory and/or can be transmitted out of the devices 10, 20.
Turning again to the drawings,
As shown in
The communication module 120 may include one or more standard communication ports including but not limited to communication ports configured to operate according to RS-232, RS-482, IEEE 802.3, IEEE 802.11, etc. In this embodiment, the firmware 104 generally comprises software instructions that when executed by processor 110 configure and control elements of the hardware platform 102. The firmware 104 may be stored in the NVM 112 and copied to and executed by the processor 110 from RAM 114. Applications 108 and the operating system 106 may also be stored in the storage memory 116 and copied to and executed by the processor 110 from RAM 114. The operating system 106 in this example includes software components like a scheduler 122 and a kernel 124. Examples of operating systems include LINUX, UCOS, WINDOWS, VXWORKS, PSOS, etc. The kernel 124 can include software functionality that provides software services and software objects to applications 108. By way of example and without limitation, software objects can include threads, queues, and/or semaphores. Applications 108 may invoke functionality in the kernel 124 to create these software objects. Applications 108 include applications 108-1 . . . 108-N. Each application 108-1 . . . 108-N may be configured to operate within the context of a corresponding thread. The scheduler 122 determines which thread and consequently which one of applications 108-1 . . . 108-N will be executed by the processor 110. Threads may be assigned different priority when they are created. The scheduler 122 generally causes the processor 110 to first execute a thread with a higher priority. Threads and their corresponding applications when executed by the processor 110 may invoke common software routines, such as software instructions to transmit and receive data from the input/output ports and to read/write to the NVM 112.
The event code generation unit 118 may receive events from the hardware platform 102, firmware 104, operating system 106, and applications 108. In response to receiving an event, the event code generation unit 118 may generate a time stamp and associate the generated time stamp with the event in a log. For example, a log entry can include the generated time stamp, an identifier of the corresponding event, and status information of the hardware platform 102, firmware 104, operating system 106, and/or applications 108. Accordingly, in one embodiment, the log includes not only an indication of the event but also the time stamp. In some embodiments, the first device 10 can transmit the log (e.g., with the generated time stamp and information about the corresponding event associated with the generated time stamp) to the event tracing unit 20 via the test port 121 or via the communication module 120.
Turning again to the drawings,
The software event capture register 210 may be configured to receive events from firmware sources including firmware capture registers 212 and application software capture registers 214. Firmware capture registers 212 may receive events generated by firmware 104 of
In some embodiments, each or some of the firmware capture registers 212, the application software capture registers 214, the storage interface 204, the host-interface 206, and the hardware cores 208 may include additional logic elements that may generate the events based on events received from their respective sources. For example, events generated by the application software capture registers 214 in response to receiving an event from application 108-1 of
The generated event codes may be transmitted via external first in first out (FIFO) port 218. Separately, the generated event codes may be stored in cyclic buffer FIFO 220. A high speed bus master like advanced high-performance bus (AHB) master 222 may transmit the generated event codes. The event tracing unit 20 may receive the event codes via either a bus connected to external FIFO port 218 or via a bus configured to operate in accordance with the protocol used by the AHB master 222. While the event code generating unit 118 may be a hardware module, such as an application specific integrated circuit (ASIC) or other integrated circuit as described above, in other embodiments it is contemplated that it may be implemented in firmware with a dedicated or shared processor.
Returning to the drawings,
In this embodiment, the processor 510 is adapted to receive an indication from the communication interface 502 when the communication interface 502 stores an event code in the buffer 504. For example, the communication interface 502 may generate an interrupt when an event code is stored in the buffer 504. The communication interface 504 may also provide the processor 510 with a reference to a memory location in the buffer 504 where the event code is stored, so that the processor 510 can retrieve the event code from the buffer 504.
In operation, the first device 10 is configured to detect internal events and make a record of them with the associated time stamp that was generated when the event occurred. Examples of events internal to the first device 10 include, but are not limited to, a hardware interrupt triggered by the reception of data, completion of a read and/or write operation to memory, the reading and/or writing of a particular variable in software, the setting of memory access triggers, bus traffic, and when difficult processing tasks are performed (e.g., simultaneous decoding of data). When an event is detected, an identification of the event and the time stamp that was generated by the time stamp generator 15 of the first device 10 when that event was detected are stored (e.g., in a log in one of the memory units in the first device 10). The identification of the event can include, for example, an event code that identifies the event and/or information about the source of the event and the operational status of the source.
The event code and time stamp can be sent to the event tracing unit (the second device) 20 for future analysis or transmission. For example, the processor 510 in the event tracing unit 20 can be configured to use the event code and time stamp to trace the operation of the hardware and software in the first device 10 to, for example, generate a chronological depiction of the occurrence of the events and display them on the display 512 and/or transmit them to another device.
In addition to the event codes and time stamps received from the first device 10, the event tracing unit 20 may be detecting its own events and recording them along with its own generated time stamps. For example, the event tracing unit 20 may be monitoring how much power the first device 10 is using over time (e.g., when the event tracing unit 20 is connected to the power supply of the first device 10); monitoring current, voltage, and/or temperature disturbances; and/or keeping a protocol log of what commands are being send to the first device 10 and what responses are received from the first device 10. So, the event tracing unit 20 can store two logs—one received from the first device 10 (of events it detects and time stamps it generates) and one produced by the event tracing log 20 (of events it detects and time stamps it generates).
It is often desired for testing and debugging purposes to correlate these logs to determine what events were occurring in the first device 10 when certain other events were detected by the event tracing unit 20 in order to determine a relationship between the events. For example, if the event tracing unit 20 detects a high power usage by the first device 10, it may be useful to know what activity was occurring at that time, so that a relationship between those events can be seen. If the time stamps share a common time domain, the generated time stamps can be used to correlate the logs. However, as discussed above, the time stamps generated by the time stamp generators 15, 25 of the two devices 10, 20 are in different time domains, meaning that they are each tracking time relative to different time origins. For example, each time stamp generators 15, 25 may be tracking time with respect to its own arbitrary initial point in time, such as when each device 10, 20 was powered on. Because the time stamps in the two logs are not correlated with each other, and the time stamps, by themselves, probably cannot be used to correlate the events in the logs. Ad hoc comparison methods can be used to correlate the time stamps by checking for patterns of external activity. However, ad hoc comparison methods are often a non-trivial undertaking and may provide limited timing accuracy.
The following embodiments can be used to correlate time stamps from different time domains. While these embodiments are helpful in correlating logs generated in the process of testing and debugging a System on Chip (SoC), it is important to note that these embodiments are not limited to System on Chip (SoC) testing. Indeed, these embodiments can be used in any situation in which time stamps are generating in two or more different time domains and need to be correlated. For example, these embodiments can be used in environments where a high level of timing accuracy is needed (e.g., because highly-accurate synchronization enables correlation of “fast” events external to a device—such as the power being consumed—with “fast” actions occurring inside the device).
In one embodiment, a time stamp synchronization event common to both the first and second devices 10, 20 is provided. Each device 10, 20 generates a time stamp in response to the time stamp synchronization event. The first device sends its generated time stamp to the second device, and the second device correlates its generated time stamp with the time stamp received from the first device. Correlating the first and second time stamps provides a relationship between the different time domains of the first and second devices 10, 20 because the first and second time stamps were both generated with respect to the time stamp synchronization event common to the first and second devices 10, 20. The second device 20 can then use this relationship to translate between the time domains and correlate various events associated with the time stamps generated by the devices 10, 20 before and after the time stamp synchronization event. It should be noted that while, in this example, the time stamp synchronization event is generated by the first device 10, in other embodiments, the time stamp synchronization event is generated by the second device 20 or by a device other than the first and second devices 10, 20. Also, instead of the second device 20 doing the correlating of the time domains, the correlating can be done by the first device or by a device other than the first and second devices 10, 20.
By allowing accurate synchronization of various time stamp domains, these embodiments can be used to compare logs, for example. With reference to the testing and debugging example given above, this embodiment provides a highly-accurate correlation between time stamp domains that can be used to correlate events external to the first device 10 (e.g., power being consumed) with actions going on in the first device 10. By taking the time stamp at the same time in both domains and doing the correlation later on, this embodiment provides an accurate way of seeing the time stamps in both domains and cancelling out for the delays of bringing the events out of the system.
Any suitable time stamp synchronization event can be used. The following paragraphs provide one example implementation illustrating the operation of an embodiment. It is important to note that this is merely an example and other implementations can be used. Accordingly, the particular type of time stamp synchronization event and how it is used in this example should not be read into the claims unless expressly recited therein.
As mentioned above, this embodiment uses an event that is externally visible to both sides of the interface between the first and second devices 10, 20 to generate a synchronization point between the timers running on different time domains in the devices 10, 20. In one particular embodiment, a bus (e.g., an Advanced Trace Bus (ATB) (“trace bus”)) allows the first device 10 to transmit information to the second device 20 (the bus can enable one-way or two-way communication). For example, when the first device 10 takes the form of a System on Chip (SoC) and the second device 20 takes the form of test equipment, the trace bus can send the event codes and time stamps generated by the SoC to the test equipment. In one particular implementation, information from the first device 10 is conveyed via the trace bus using a time division multiplex (TDM) frame that has a header (e.g., 8 bits) and a payload (e.g., 64 bits). In this example, after every N frames (or using some other periodic or non-periodic timing), a special header is used that is recognized by the second device 20 as being the time stamp synchronization event. The second device 20 generates a time stamp in response to receiving the special header. The second device 20 also receives a time stamp from the first device 10 that it generated in response to the special header. With this pair of time stamps, the second device 20 can correlate the two time stamps and use that relationship to “translate” between the two time domains.
With reference again to the drawings,
In this example, the Advanced Trace Bus (ATB) uses a time division multiplexed (TDM) frame with a header and payload. The ATB header is the mechanism used to provide the common synchronization event in this example. Specifically, the common synchronization event takes the form of a special ATB frame header in this example.
When the ATB generates the special sync header (act 610), it is detected by a signal sniffer 17 in the first device 10 and the ATB 26 in the second device 20. When the signal sniffer 17 detects the special sync header, the signal sniffer 17 generates a signal sniffer event, which triggers a system event tracer (SET) event 18 (act 630). The SET event causes the first device 10 to generate an event code 28 for the synch event and associates a time stamp with that event code, both of which are provided to the ATB 16 (e.g., in the payload of the ATB TDM frame), for transmission to the second device 20, which can store the information in an event log 30, which can be internal or external to the second device 20 (act 650).
In the meantime, when the ATB 26 of the second device 20 receives the special synch header from the first device 10, an ATB synch detector 27 detects the special header and responds by storing a time stamp generated at that time by the second device 20 in a storage device 29 in the second device 20 (act 620). When the second device 20 sees that it has received an event code from the first device 10 that indicates a synchronization event, the second device 20 issues a command (act 660) to attach the time stamp generated by the second device 20 when the special header was received to the received event code from the first device 10. The time stamp from the second device 20 is then also stored in the event log 20 (act 670). The monitor detection of the ATB sync with its time stamp and the first device's event code with its time stamp can both be written into the log consecutively.
The result of these operations is that the event log 30 stores an event code indicates a time synchronization event, along with time stamps from the first and second devices 10, 20 (the event code can be used to allow the second device 20 to distinguish the synchronization event in the log from other entries received by the second device 20). As these time stamps were associated with the same, common time synchronization event, the offset between these time stamps can be used to determine a relationship between the two time domains. This can be used to align various events detected by the first device 10 with other various events detected by the second device 20 in order to find a possible connection between those events. Successive sets of time stamps can also be used to understand the frequency relationship between the clocks in the first and second devices 10, 20 (e.g., the first device 10 may generate two clock pulses for every one clock pulse generated by the second device 20).
It should be noted that, in this embodiment, the determined relationship between the first and second time domains is known by the second device 20 and is not transmitted to the first device 10, as the first device 10 does not need to know that relationship in this example. However, in other embodiments, the second device 20 can provide the relationship to the first device 10 (or to another device).
There are many alternatives that can be used with these embodiments. For example, one alternate embodiment can be used to synchronize a plurality of logs with different clock domains and time stamp representations, such as, for example, kernel message (kmsg) and logcat logs in the Android platform and a log in a storage device. In general, kmsg and logcat logs are typically the main logs in the Android platform and are used by developers for validation and to determine what happened in the system. It is difficult to correlate these logs in time because they have different clock sources. More specifically, the kmsg log uses a local clock (or “shed” clock) that may stop during a suspend/sleep operation. In contrast, the logcat log uses a real-time clock (a “wall clock”) (CLOCK_REALTIME). A storage device typically uses yet another clock domain.
It is also difficult to correlate these logs in time because they have different representation formats. For example, the kmsg log is formatted in microseconds since platform boot (e.g., “6,33304,2739220275,-;[pid:175,cpu0,kworker/0:1H]watchdog watchdog0:watchdog”), whereas the logcat log is in a real-time (or “wall-clock time”) format (e.g., 12-14 08:44:33.122952 1919 2195 E LogCollectService: illegal eventid”). Storage device logs are typically in yet another format.
Several mechanisms can be used to attempt to correlate kmsg and logcat logs; however, they each have their own disadvantages. For example, in one approach, logcat and kmsg logs are merged into one log in real time by redirecting the logcat messages into the kmsg log, or vice versa. However, this may negatively impact the combined log's readability when the device needs to read only the kmsg log or only the logcat log (typically, only the kmsg entries or the logcat entries need to be analyzed during log analysis for a sequence of events).
Another approach is to attach a “wall time” time stamp of the format and clock source used in the logcat log to every kmsg message in the Linux kernel code. However, this approach can be intrusive and may only be limited to kmsg and logcat logs. This may also not be a practical solution for synchronizing firmware trace (ftrace) log and logcat logs due to the ftrace event size, which can negatively impact the overall system performance. Ftrace is an internal tracer designed to help kernel developers find what is going on inside the kernel. It is yet another Linux log with its own timestamp not synchronized to the logcat. The ftrace log is much more verbose and can generate much more entries per second than kmsg. Adding a common time stamp to every entry will increase the entry size and as consequence, the ftrace log size.
Another approach is to have a dedicated service running on the Android platform that retrieves messages from the logs, attaches a common time stamp, and forwards them to a collecting device, such as a personal computer (PC, laptop, desktop, workstation, dedicated appliance, or any other computing device). While this approach can provide good accuracy, it can change platform behavior, have an impact on the platform's performance, be intrusive, and have bad scalability.
In addition to the disadvantages noted above, these approaches may be even less practical when there is a necessity to synchronize these logs to a log of a storage device (e.g., a firmware trace log retrieved directly from the storage device) that is recorded in a different time domain. Further, since firmware traces are typically retrieved by dedicated hardware external to the Android platform, they may not be able to be intermixed with Android logs in a single file. And due to different time domains, their time stamps cannot be easily correlated to the time stamps of the logcat and kmsg logs. The following embodiments provide a solution to this problem
Turning now to
In this embodiment, computer 1 comprises a clock operating in a first clock domain (clock domain 1) and a collector (collector 1), and computer 2 comprises a clock operating in a second clock domain (clock domain 2) and a collector (collector 2). Although not shown in
In this embodiment, the host 710 comprises a processor that generates a plurality of logs (here, kmsg and logcat logs, although additional or other logs can be used 720, a clock operating in a third clock domain (clock domain 3), a storage device 730 with a log (here, a firmware log), and a calibration time stamp daemon (CTSD) 740.
The clocks in the host 710 and the two computers can operate in different time domains, and a time server (here, a Network Time Protocol (NTP) server) 750 can be used to synchronize the three clock domains (e.g., the Android platform and computers can belong to different time domains, and the NTP server 750 can be used to keep the domains in sync). However, even with such synchronization, there can be a skew (non-synchronization) in one or more of the clocks (e.g., due to different crystals used in the clocks, due to network latency variation, due to the location of the server 750, etc.). Such deviation can make it difficult or impossible to correlate the logs.
In this embodiment, to measure and account for possible skew, the host 710 uses the CTSD 740 to periodically inject (e.g., once a second, but any suitable time value can be used) a calibration message with a time stamp (TS1) into each log entry created at that time (e.g., the kmsg, logcat, and storage device log entries). This is shown as acts 801, 802, and 803 in the timing diagram of
The logs are then sent to different computing devices for analysis. In this example, the kmsg and logcat logs are sent to computer 1 (acts 804 and 805). Collector 1 in computer 1 retrieves entries from the kmsg and logcat logs, and a clock running on clock domain 1 places time stamps (TS2, TS3) on each entry retrieved from the log (acts 807 and 808). Similarly, the host 710 sends the storage device's firmware log to computer 2 (act 806), and collector 2 in computer 2 retrieves entries from the firmware log, and a clock running on clock domain 2 places a time stamp (TS4) on an entry retrieved from the log (act 810). The log entries, including the dual-time-stamped log entries of the calibration messages, are then sent from computers 1 and 2 to an external database 760. In one embodiment, not all log entries are dual-time stamped—only log entries of calibration messages (“calibration log entries”), as creating a dual time stamp for all log entries can result in an intrusive method of synchronization.
As noted above, even though all three clocks are synchronized by the NTP 750, there can be skew among the clocks, varying as time passes. For example, in one embodiment, NTP synchronization accuracy ranges between a few microseconds for wired networks to over 10 microseconds for wireless networks. This skew can reach seconds when the NTP service of one of the computers participating in the given setup dies or is restarted. This can be even worse when the test setup incorporates a computer running Windows and Linux, where NTP clients are implemented differently. Leap seconds and daylight savings time can cause even greater jumps between the clocks. Also, due to the differences in the logging mechanisms of logcat and kmsg logs and the way the messages from the logs are retrieved, the skew between the time stamps applied by computer 1 to the log entries can be a jitter in tens of milliseconds range (TS2 and TS3).
The common time stamp provided to all calibration log entries by the host 710 can be used, along with the secondary time stamp application by the computers 1, 2 to measure and account for the skew. For example, by having a common time stamp (T1), various skews between different clock domains can be measured (e.g., TS1-TS2, TS1-TS3, TS1-TS4) to provide an alert as to an unacceptable level of clock skew deviation. As another example, the common time stamp can be corrected with a value calculated as a difference between the time stamps of computers 1 and 2 (TS2, TS3, and TS4) and the common time stamp (TS1) with consequent interpolation within the calibration time stamp period. Further, the common time stamp allows for manually correlating between fragments of the different logs in a setup mode running without the common time stamp attached at the collector side. More generally, these embodiments can provide an indication of skew between all three clocks: SKEW1=TS2-TS1—retrieval_time_average; SKEW2=TS4-TS1—retrieval_time_average and SKEW3=TS4-TS2. In one particular embodiment, the interest is in the skew between TS2, TS3, and TS4 timestamps in the logs.
There are many advantages associated with these embodiments. For example, these embodiments can be used to synchronize logs of the Android platform between themselves and with firmware traces of the storage device 730, overcoming the disadvantages discussed above with other approaches. This provides a non-intrusive, lightweight method of log synchronization, which keeps Android platform behavior and performance intact. As another example, these embodiments can be used to shorten resolution time of storage system firmware bugs originated in the Android platform, positively impacting time to market of storage system products. Additionally, while some of the embodiments discussed earlier provide time stamp synchronization at the hardware level, they rely on bus timing, whereas these embodiments can be implemented in software running above the Android platform.
It also should be noted that the embodiments discussed above may be implemented exclusively in hardware or firmware or in any combination thereof. Additionally, in embodiments where hardware is utilized, the hardware may include field programmable gate arrays (FPGAs), application specification integrated circuits (ASICs), general purpose digital logic or any combination thereof. Separately, where firmware or software is utilized to implement the embodiments discussed herein, machine instructions corresponding to the firmware or software may be stored in a suitable non-transitory storage medium that may be read by a processor.
Further, each of the methods described herein may be encoded in a computer-readable storage medium (e.g., a computer memory), programmed within a device (e.g., one or more circuits or processors), or may be processed by a controller or a computer. If the processes are performed by software, the software may reside in a local or distributed memory resident to or interfaced to a storage device, a communication interface, or non-volatile or volatile memory in communication with a transmitter. The memory may include an ordered listing of executable instructions for implementing logic. Logic or any system element described may be implemented through optic circuitry, digital circuitry, through source code, through analog circuitry, or through an analog source, such as through an electrical, audio, or video signal. The software may be embodied in any computer-readable or signal-bearing medium, for use by, or in connection with an instruction executable system, apparatus, or device. Such a system may include a computer-based system, a processor-containing system, or another system that may selectively fetch instructions from an instruction executable system, apparatus, or device that may also execute instructions.
Finally, as mentioned above, any suitable type of memory can be used. Semiconductor memory devices include volatile memory devices, such as dynamic random access memory (“DRAM”) or static random access memory (“SRAM”) devices, non-volatile memory devices, such as resistive random access memory (“ReRAM”), electrically erasable programmable read only memory (“EEPROM”), flash memory (which can also be considered a subset of EEPROM), ferroelectric random access memory (“FRAM”), and magnetoresistive random access memory (“MRAM”), phase-change memory (“PCM”) and other semiconductor elements capable of storing information. Each type of memory device may have different configurations. For example, flash memory devices may be configured in a NAND or a NOR configuration.
The memory devices can be formed from passive and/or active elements, in any combinations. By way of non-limiting example, passive semiconductor memory elements include ReRAM device elements, which in some embodiments include a resistivity switching storage element, such as an anti-fuse, phase change material, etc., and optionally a steering element, such as a diode, etc. Further by way of non-limiting example, active semiconductor memory elements include EEPROM and flash memory device elements, which in some embodiments include elements containing a charge storage region, such as a floating gate, conductive nanoparticles, or a charge storage dielectric material.
Multiple memory elements may be configured so that they are connected in series or so that each element is individually accessible. By way of non-limiting example, flash memory devices in a NAND configuration (NAND memory) typically contain memory elements connected in series. A NAND memory array may be configured so that the array is composed of multiple strings of memory in which a string is composed of multiple memory elements sharing a single bit line and accessed as a group. Alternatively, memory elements may be configured so that each element is individually accessible, e.g., a NOR memory array. NAND and NOR memory configurations are examples, and memory elements may be otherwise configured.
The semiconductor memory elements located within and/or over a substrate may be arranged in two or three dimensions, such as a two dimensional memory structure or a three dimensional memory structure.
In a two dimensional memory structure, the semiconductor memory elements are arranged in a single plane or a single memory device level. Typically, in a two dimensional memory structure, memory elements are arranged in a plane (e.g., in an x-z direction plane) which extends substantially parallel to a major surface of a substrate that supports the memory elements. The substrate may be a wafer over or in which the layer of the memory elements are formed or it may be a carrier substrate which is attached to the memory elements after they are formed. As a non-limiting example, the substrate may include a semiconductor such as silicon.
The memory elements may be arranged in the single memory device level in an ordered array, such as in a plurality of rows and/or columns. However, the memory elements may be arrayed in non-regular or non-orthogonal configurations. The memory elements may each have two or more electrodes or contact lines, such as bit lines and wordlines.
A three dimensional memory array is arranged so that memory elements occupy multiple planes or multiple memory device levels, thereby forming a structure in three dimensions (i.e., in the x, y and z directions, where the y direction is substantially perpendicular and the x and z directions are substantially parallel to the major surface of the substrate).
As a non-limiting example, a three dimensional memory structure may be vertically arranged as a stack of multiple two dimensional memory device levels. As another non-limiting example, a three dimensional memory array may be arranged as multiple vertical columns (e.g., columns extending substantially perpendicular to the major surface of the substrate, i.e., in the y direction) with each column having multiple memory elements in each column. The columns may be arranged in a two dimensional configuration, e.g., in an x-z plane, resulting in a three dimensional arrangement of memory elements with elements on multiple vertically stacked memory planes. Other configurations of memory elements in three dimensions can also constitute a three dimensional memory array.
By way of non-limiting example, in a three dimensional NAND memory array, the memory elements may be coupled together to form a NAND string within a single horizontal (e.g., x-z) memory device levels. Alternatively, the memory elements may be coupled together to form a vertical NAND string that traverses across multiple horizontal memory device levels. Other three dimensional configurations can be envisioned wherein some NAND strings contain memory elements in a single memory level while other strings contain memory elements which span through multiple memory levels. Three dimensional memory arrays may also be designed in a NOR configuration and in a ReRAM configuration.
Typically, in a monolithic three dimensional memory array, one or more memory device levels are formed above a single substrate. Optionally, the monolithic three dimensional memory array may also have one or more memory layers at least partially within the single substrate. As a non-limiting example, the substrate may include a semiconductor such as silicon. In a monolithic three dimensional array, the layers constituting each memory device level of the array are typically formed on the layers of the underlying memory device levels of the array. However, layers of adjacent memory device levels of a monolithic three dimensional memory array may be shared or have intervening layers between memory device levels.
Then again, two dimensional arrays may be formed separately and then packaged together to form a non-monolithic memory device having multiple layers of memory. For example, non-monolithic stacked memories can be constructed by forming memory levels on separate substrates and then stacking the memory levels atop each other. The substrates may be thinned or removed from the memory device levels before stacking, but as the memory device levels are initially formed over separate substrates, the resulting memory arrays are not monolithic three dimensional memory arrays. Further, multiple two dimensional memory arrays or three dimensional memory arrays (monolithic or non-monolithic) may be formed on separate chips and then packaged together to form a stacked-chip memory device.
Associated circuitry is typically required for operation of the memory elements and for communication with the memory elements. As non-limiting examples, memory devices may have circuitry used for controlling and driving memory elements to accomplish functions such as programming and reading. This associated circuitry may be on the same substrate as the memory elements and/or on a separate substrate. For example, a controller for memory read-write operations may be located on a separate controller chip and/or on the same substrate as the memory elements.
One of skill in the art will recognize that this invention is not limited to the two dimensional and three dimensional structures described but cover all relevant memory structures within the spirit and scope of the invention as described herein and as understood by one of skill in the art.
It is intended that the foregoing detailed description be understood as an illustration of selected forms that the invention can take and not as a definition of the invention. It is only the following claims, including all equivalents, that are intended to define the scope of the claimed invention. Finally, it should be noted that any aspect of any of the embodiments described herein can be used alone or in combination with one another.
This application is a continuation-in-part of U.S. patent application Ser. No. 15/820,147, filed Nov. 21, 2017, which is hereby incorporated by reference herein.
Number | Name | Date | Kind |
---|---|---|---|
6898213 | Shimelmitz | May 2005 | B1 |
7251748 | Liberty | Jul 2007 | B2 |
7929523 | Shenfield et al. | Apr 2011 | B2 |
9400734 | Aharonov et al. | Jul 2016 | B2 |
9742796 | Salsamendi | Aug 2017 | B1 |
10387226 | Brief | Aug 2019 | B2 |
20030002540 | Eerenberg | Jan 2003 | A1 |
20030158706 | Eidson | Aug 2003 | A1 |
20050021708 | Raghuraman et al. | Jan 2005 | A1 |
20100042632 | Johnson | Feb 2010 | A1 |
20100107143 | Emberling | Apr 2010 | A1 |
20130004142 | Grab | Jan 2013 | A1 |
20140355629 | Ito | Dec 2014 | A1 |
20150022262 | Du | Jan 2015 | A1 |
20150082325 | Aharonov et al. | Mar 2015 | A1 |
20170060479 | Hong | Mar 2017 | A1 |
20170186297 | Brenner | Jun 2017 | A1 |
20170289720 | Tsukagoshi | Oct 2017 | A1 |
20180006866 | Trainin | Jan 2018 | A1 |
20180224887 | Pitigoi-Aron | Aug 2018 | A1 |
20180254825 | Speidel | Sep 2018 | A1 |
20190020432 | Chaudhari | Jan 2019 | A1 |
Entry |
---|
U.S. Appl. No. 15/811,123, filed Nov. 13, 2017, Brief et al. |
Office Action dated Jan. 7, 2019 in U.S. Appl. No. 15/811,123. |
Office Action dated Apr. 24, 2019 in U.S. Appl. No. 15/820,147. |
Stack Overflow; “Android—How to sync Kernel time and logcat time?”; downloaded from the Internet on Apr. 4, 2018 at https://stackoverflow.com/questions/6329648/how-to-sync-kernel-time-and-logcat-time; 3 pages. |
Stack Exchange; “Logging—How can I view and examine the Android log?”; downloaded from the Internet on Apr. 4, 2018 at https://android.stackexchange.com/questions/14430/how-can-i-view-and-examine-the-android-log; 5 pages. |
Stack Overflow; “Is Android using NTP to sync time?”; downloaded from the Internet on Apr. 4, 2018 at https://stackoverflow.com/questions/14381005/is-android-using-ntp-to-sync-time; 3 pages. |
Stack Overflow; “Android—Taking logcat and kernel logs simultaneously”; downloaded from the Internet on Apr. 4, 2018 at https://stackoverflow.com/questions/22652605/taking-logcat-and-kernel-logs-simultaneously; 1 page. |
U.S. Appl. No. 15/820,147 entitled “System and Method for Time Stamp Synchronization,” filed Nov. 21, 2017, 34 pages. |
Office Action in U.S. Appl. No. 15/820,147, dated Jan. 14, 2020, 9 pages. |
Office Action in U.S. Appl. No. 15/820,147, dated Aug. 30, 2019, 22 pages. |
Number | Date | Country | |
---|---|---|---|
20190155327 A1 | May 2019 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 15820147 | Nov 2017 | US |
Child | 16012236 | US |