This invention relates to extender cards, and more particularly to memory module extender cards with visual indicators for debugging.
Many computer systems such as personal computers (PCs) use memory modules as their main memory. Memory modules may be tested using PC-motherboard-based testers. Servers may use many memory modules.
When a memory modules fails on a server, it can be difficult to determine which of the many memory modules has failed. A software tool such as a memory diagnostics program can be executed on the server to write and read locations in the memory. However, these memory diagnostic programs access virtual or logical addresses, rather than the physical addresses of the memory modules.
The operating system (OS) running on the server (in conjunction with BIOS) may remap logical addresses while the processor and chips such as a north bridge chip translate logical addresses from the processor to physical addresses. While the address translation is deterministic, it may not be known to the end user, since manufacturers often keep address translation as a trade secret.
When a memory diagnostics program is executed on a processor or Central Processing Unit (CPU), a sequence of logical addresses are written and read back. Mismatched data read back indicates a faulty memory location within the logical address space. Since the mapping from the logical address space to the physical address space and the memory modules is not known to the user, it cannot be readily determined which memory module is malfunctioning despite the memory diagnostic program having located the error by its logical address. A technician cannot easily determine which of the many memory modules to replace despite running the memory diagnostic program and may have to replace memory modules one by one until the problem is fixed.
What is desired is a diagnostic method that can identify which memory module is malfunctioning. An extender card with a visual indicator is desired to show which memory module is malfunctioning when a memory diagnostic program is executed.
The present invention relates to an improvement in memory module diagnostics. The following description is presented to enable one of ordinary skill in the art to make and use the invention as provided in the context of a particular application and its requirements. Various modifications to the preferred embodiment will be apparent to those with skill in the art, and the general principles defined herein may be applied to other embodiments. Therefore, the present invention is not intended to be limited to the particular embodiments shown and described, but is to be accorded the widest scope consistent with the principles and novel features herein disclosed.
Since operating system 102 controls the execution of test program 104, when test program 104 is running, some instructions from operating system 102 may be executed at nearly the same time. For example, operating system 102 may allow test program 104 to execute, but then may pause test program 104 so that operating system 102 may execute a few instructions to update a clock or a timer, or to check the status of other programs. Then operating system 102 allows test program 104 to resume execution. Thus some instructions from operating system 102 can be executed when test program 104 is running.
This makes diagnostics difficult, because test program 104 could be set to loop on a particular memory address that is failing, but some other memory addresses are accessed during this loop when operating system 102 performs some background operation. Thus the wrong logical memory address and ultimately the wrong memory module could be pointed to by OS accesses during the test program loop.
Extender card 24 is a printed-circuit board (PCB) or other substrate that has test socket 22 mounted on its upper edge, and has metal fingers or contact pads along its bottom edge. Metal wiring traces on extender card 24 connect corresponding signals on the lower-edge contact pads to metal pads in test socket 22, thus passing signals through. However, some signals are also routed to intercepting decoder chip 30 on extender card 24. This routing allows intercepting decoder chip 30 to observe or sniff address and chip-select signal sent from PC motherboard 28 to memory module 10.
When intercepting decoder chip 30 detects that DRAM chips 18 on memory module 10 inserted into test socket 22 are being accessed, such as by detecting that PC motherboard 28 has activated the chip select signal to memory module socket 26, intercepting decoder chip 30 energizes visual indicator 60 on extender card 24. The light from visual indicator 60 is visible to a user, indicating that memory module 10 is being accessed.
PC motherboard 28 is a larger PCB that has chips, sockets, and other components mounted thereon, such as chip 32 and expansion sockets 36 which have expansion cards 34 plugged in. Expansion cards 34 can be Peripheral Component Interconnect (PCI), Peripheral Components Interconnect Express (PCIE), AT-bus, or other expansion cards. Chip 32 can be the main microprocessor, chip set, cache memory, or other chips.
Memory module socket 26 is one of several sockets designed to fit memory module 10 or other memory modules. Memory module socket 26 is mounted to the PCB substrate of PC motherboard 28, and fits the contact pads on the bottom side of extender card 24, or the contact pads on memory module 10.
Address signals 48, data signals 46, and control signals 44 from DRAM controller 38 on PC motherboard 28 are passed through extender card 24 and sockets 22, 26 to reach DRAM chips 18 on memory module 10. When DRAM chips 18 are synchronous DRAMs, control signals 44 can include a clock signal or strobes.
Intercepting decoder chip 30 examines the chip select signal from control signals 44 and activates visual indicator 60 when the chip select signal is active. DRAM controller 38 decodes the address from the CPU on motherboard 28, such as a logical address, and determines which of the possible many memory modules to activate. A different chip select may be generated for each memory module, or additional decoding may be performed. When the chip select line to memory module socket 26 is activated, memory module 10 is typically accessed while other memory modules in other sockets are not accessed.
Intercepting decoder chip 30 can latch or extend the pulse width of the chip select signal so that visual indicator 60 remains on for longer than the duration of the access. For example, a trigger circuit could latch the chip select signal using a clock, then keep the drive signal to visual indicator 60 active for some number of clocks, such as 10 clocks.
Intercepting decoder chip 30 may decode some of the address signal 48 before activating visual indicator 60. Intercepting decoder chip 30 may also latch in the address, or a part of the address, and display this latched address as a hexadecimal number on a liquid crystal display (LCD) on extender card 24 (not shown).
Test program 50 has instructions to write and read back many logical address locations. When these instructions are executed on the CPU on motherboard 28, logical addresses are sent to DRAM controller 38, which re-maps these logical addresses into physical addresses that include address signals 48 and control signals 44 (
When DRAM controller 38 reads or writes to a physical address in memory module 10, it activates the chip select (CS) to memory module socket 26. Intercepting decoder chip 30 responds to the chip select signal by activating or illuminating visual indicator 60.
DRAM controller 38 can be configured for the memory size, type, and timing of memory module 10. Addresses from the microprocessor can be routed to different row and column address lines to DRAM chips 18 for different memory sizes. The number of clock cycles between control signals sent from DRAM controller 38 to DRAM chips 18 can be adjusted to meet the timing parameters in the configuration. BIOS can verify the configuration by writing and reading back data from locations in the configured memory of DRAM chips 18.
More extensive test programs 50 can be executed that write and read each location in DRAM chips 18 using a variety of test patterns such as walking ones and zeros, checkerboard, etc. These test patterns are executed on the microprocessor on PC motherboard 28 from test program 50, although some test programs may also exist in the BIOS.
Since extender card 24 passes through signals from DRAM controller 38 to DRAM chips 18, DRAM chips 18 can be tested as if memory module 10 was plugged directly into socket 26 on PC motherboard 28.
Should testing of DRAM chips 18 reveal a fault, test program 50 can be set to continuously or repeatedly loop through the faulty memory location, writing and reading the bad logical address. When there are many memory modules and sockets on PC motherboard 28, each with an extender card 24, the extender card 24 with its visual indicator 60 illuminated has the memory module 10 with the bad memory location. Thus the memory module having the physical address that maps to the logical address being looped to in test program 50 is identified by illumination of visual indicator 60 on extender card 24 that the faulty memory module 10 is plugged into.
The inventors realize that the majority of the memory accesses should be from test program 50 while a smaller minority of accesses are from the operating system. The inventors have devised a filtering circuit that filters out the fewer accesses that are presumed to be caused by the operating system, allowing the greater number of accesses due to test program 50 to activate visual indicator 60.
A minimum threshold or accesses can be set. The number of accesses per time period by the operating system should fall below this minimum threshold, while the number of accesses per time period by test program 50 should be above this minimum threshold.
Timer 62 is clocked by the clock from DRAM controller 38 to memory module 10 that is in control signals 44 (
Counter 64 is triggered to increment when a valid memory access occurs to the memory module inserted into test socket 22 on extender card 24 containing this intercepting decoder chip 30. For example, when RAS or CAS is activated and the chip select CS is also activated, counter 64 is triggered and increments its count value. The address may also be decoded to generate the trigger, or the trigger may be qualified to only trigger on a write when WE is active. The trigger could also be qualified by RAS, and other latching logic could latch in the row address from address signals 48 (
The count value from counter 64 is compared to a minimum threshold by decision logic 68. When the count value from counter 64 exceeds the minimum threshold, decision logic 68 drives an activation signal to visual indicator 60, causing 60 to be illuminated.
Intercepting decoder chip 30 could be programmable so that the test program or another diagnostic program could determine whether the row or the column address is displayed on LCD 61. LCD 61 can be a small liquid crystal display (LCD), or could be some other kind of display. Additional display driver chips could be present, or could be part of a component module for LCD 61. The display driver logic could also be integrated into intercepting decoder chip 30.
A technician could turn off a faulty server, remove all the memory modules, then insert an extender card 24 into each memory module socket 26, then plug memory modules 10 into test sockets 22 in each of the extender cards 24. The server can be rebooted and test program 50 executed. When a failing address is located by test program 50, then test program 50 can jump to a looping routine that continuously writes and reads just the one faulty memory location. Visual indicator 60 lights up for the memory module being accessed by that faulty memory location, since most accesses are to the physical address corresponding to the logical address being accessed by test program 50. Operating system accesses may occur, but these can be filtered out so that they do not illuminate visual indicator 60, such as by using the logic of
The technician can then remove the memory module plugged into extender card 24 with the illuminated visual indicator 60 and insert a new memory module into extender card 24. The test program can be re-run, this time finding no faults. The server can then be restored to server once the technician removes all of the extender cards 24 and replaced the good memory modules into memory module sockets 26. Alternately, the technician could leave extender cards 24 in the server for use at a later time.
Alternate Embodiments
Several other embodiments are contemplated by the inventors. For example visual indicator 60 could be a light-emitting diode (LED), a liquid crystal display (LCD), or some other visual indicators. Visual indicator 60 could change colors when triggered, or could flash or display continuously. Other forms of an output indication such as audio or a log file that is transmitted over a computer network could be substituted for visual indicator 60.
Some memory modules may contain multiple banks or ranks. Additional chip select signals can be provided by the motherboard to access the different ranks in a memory module. The chip selects may be encoded with the rank information. Intercepting decoder chip 30 can receive several CS signals and perform additional detection of the different ranks of the inserted memory module.
The memory module or extender card 24 may contain additional components, such as passive capacitors and resistors, and active components such as buffer chips, and registers for buffering control, address, or data lines.
The PC motherboard can be a standard motherboard, or can be a modified board, such as one having the socket for the extender card reverse-mounted on the solder side rather than the component side of the motherboard substrate. The PC motherboard can be one of several in a larger test system, such as in a robotic test system. A handler can automatically insert and remove the memory modules being tested, or a robotic arm or human operator can handle the memory modules under test.
More than one of the memory module slots on the PC motherboard may be loaded with an extender card, allowing two or more memory modules to be tested at the same time by the same PC motherboard. Each memory module socket or slot may have a different value of device-address lines A2, A1, A0, so each slot can be accessed separately. Other device-address lines could be intercepted. Intercepting decoder chip 30 could be an Application-Specific Integrated Circuit (ASIC), a programmable logic chip, a custom logic chip, or some other technology.
Other filtering circuits may be used to filter out the OS memory accesses in
The background of the invention section may contain background information about the problem or environment of the invention rather than describe prior art by others. Thus inclusion of material in the background section is not an admission of prior art by the Applicant.
Any methods or processes described herein are machine-implemented or computer-implemented and are intended to be performed by machine, computer, or other device and are not intended to be performed solely by humans without such machine assistance. Tangible results generated may include reports or other machine-generated displays on display devices such as computer monitors, projection devices, audio-generating devices, and related media devices, and may include hardcopy printouts that are also machine-generated. Computer control of other machines is another tangible result.
Any advantages and benefits described may not apply to all embodiments of the invention. When the word “means” is recited in a claim element, Applicant intends for the claim element to fall under 35 USC Sect. 112, paragraph 6. Often a label of one or more words precedes the word “means”. The word or words preceding the word “means” is a label intended to ease referencing of claim elements and is not intended to convey a structural limitation. Such means-plus-function claims are intended to cover not only the structures described herein for performing the function and their structural equivalents, but also equivalent structures. For example, although a nail and a screw have different structures, they are equivalent structures since they both perform the function of fastening. Claims that do not use the word “means” are not intended to fall under 35 USC Sect. 112, paragraph 6. Signals are typically electronic signals, but may be optical signals such as can be carried over a fiber optic line.
The foregoing description of the embodiments of the invention has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. It is intended that the scope of the invention be limited not by this detailed description, but rather by the claims appended hereto.
Number | Name | Date | Kind |
---|---|---|---|
5467314 | Miyazawa et al. | Nov 1995 | A |
6055653 | LeBlanc et al. | Apr 2000 | A |
20060123444 | Schaffer et al. | Jun 2006 | A1 |
Number | Date | Country | |
---|---|---|---|
20130151904 A1 | Jun 2013 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 12965699 | Dec 2010 | US |
Child | 13760948 | US |