US20040054867A1 - Translation lookaside buffer - Google Patents
Translation lookaside buffer Download PDFInfo
- Publication number
- US20040054867A1 US20040054867A1 US10/242,785 US24278502A US2004054867A1 US 20040054867 A1 US20040054867 A1 US 20040054867A1 US 24278502 A US24278502 A US 24278502A US 2004054867 A1 US2004054867 A1 US 2004054867A1
- Authority
- US
- United States
- Prior art keywords
- virtual address
- translation
- line
- address
- memory
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/10—Address translation
- G06F12/1027—Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/10—Address translation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/65—Details of virtual memory and virtual address translation
- G06F2212/652—Page size control
Definitions
- This invention relates to the area of translation lookaside buffers and more specifically to translation lookaside buffer architectures for rapid design cycles.
- Virtual addressing enables the system to effectively create a virtual memory space larger than an actual physical memory space.
- the process of breaking up the actual physical memory space into the virtual memory space is termed paging.
- Paging breaks up a linear address space of the physical memory space into fixed blocks called pages. Pages allow a large linear address space to be implemented with a smaller physical main memory plus cheap background memory. This configuration is referred to as “virtual memory.”
- Paging allows virtual memory to be implemented by managing memory in pages that are swapped to and from the background memory. Paging offers additional advantages, including reduced main memory fragmentation, selective memory write policies for different pages, and varying memory protection schemes for different pages.
- the presence of a paging mechanism is typically transparent to the application program.
- the size of a page is a tradeoff between flexibility and performance.
- a small page size allows finer control over the virtual memory system but it increases the overhead from paging activity. Therefore many CPUs support a mix of page sizes, e.g. a particular MIPS implementation supports any mix of 4 kB, 16 kB, 64 kB, 256 kB, 1 MB, 4 MB and 16 MB pages.
- a processor is then able to advantageously operate in the virtual address space using virtual addresses. Frequently, however, these virtual addresses must be translated into physical addresses—actual memory locations.
- One way of accomplishing this translation of virtual addresses into physical addresses is a use of translation tables that are regularly accessed and stored in main memory. Translation tables are stored in main memory because they are typically large in size. Unfortunately, regularly accessing of translation tables stored in main memory tends to slow overall system performance.
- TLB translation lookaside buffer
- a TLB is a special type of cache memory. As with other types of cache memories, a TLB is typically comprised of a relatively small amount of memory storage specially designed to be quickly accessible.
- a TLB typically incorporates both a tag array and a data array, as are provided in cache memories. Within the tag array, each tag line stores a virtual address. This tag line is then associated with a corresponding data line in the data array in which is stored a physical address translation for the virtual address.
- a processor prior to seeking a translation of a virtual address from translation tables in main memory, a processor first refers to the TLB to determine whether the physical address translation of the virtual address is presently stored in the TLB. In the event that the virtual address and corresponding physical address are stored in the TLB, the TLB provides the corresponding physical address at an output port thereof, and a time and resource-consuming access of main memory is avoided.
- a content addressable memory CAMs are parallel pattern matching circuits. In a matching mode of operation the CAM permits searching of all of its data in parallel to find a match.
- a translation lookaside buffer comprising: at least an input port for receiving a portion of a virtual address
- a random access memory a set of registers; and, synthesisable logic for determining a hash value from the received portion of the virtual address and for comparing the hash value to a stored hash value within the set of registers to determine a potential that a physical address associated with the virtual address is stored within a line within the random access memory and associated with a register, from the set of registers, within which the hash value is stored.
- a translation lookaside buffer comprising: a random access memory; a first register associated with a line in the memory; and, a hashing circuit for receiving a virtual address other than a virtual address for which a translation is presently stored in the memory, for determining a hash value and for storing the hash value in the first register; and the hashing circuit for storing the virtual address and a translation therefor in the line in memory.
- a translation lookaside buffer comprising: RAM; and, synthesisable logic for determining from a virtual address at least one potential address within the RAM in fixed relation to which to search for a physical address associated with the virtual address, the at least one potential address being other than the one and only known address within the RAM in fixed relation to which the physical address associated with the virtual address is stored.
- a method of performing a virtual address lookup function for a translation lookaside buffer including RAM and synthesisable logic including the steps of: providing a virtual address to the synthesisable logic; hashing the provided virtual address to provide a hash result;
- FIG. 1 a illustrates a prior art transistor implementation of a SRAM circuit
- FIG. 1 b illustrates a prior art transistor implementation of a CAM circuit
- FIG. 2 illustrates a prior art translation process from a virtual address (VA) to a physical address (PA);
- FIG. 3 illustrates a prior art translation from a VA to a PA when performed in conjunction with a direct mapped cache memory
- FIG. 4 a generally illustrates a translation lookaside buffer formed using synthesisable logic components and a random access memory
- FIG. 4 b illustrates a translation lookaside buffer in more detail formed from synthesisable logic components
- FIG. 4 c outlines the steps taken for operation of the TLB
- FIG. 5 illustrates a hashing circuit in more detail
- FIG. 6 illustrates a variation of the hashing circuit shown in FIG. 5.
- CAM circuits include storage circuits similar in structure to SRAM circuits. However, CAM circuits also include search circuitry offering an added benefit of a parallel search mode of operation, thus enabling searching of the contents of the CAM in parallel using hardware. When searching the CAM for a particular data value, the CAM provides a match signal upon finding a match for that data value within the CAM.
- a main difference between CAM and SRAM is that in a CAM, data is presented to the CAM representative of a virtual address and an address relating to the data is returned, whereas in a SRAM, an address is provided to the SRAM and data stored at that address is returned.
- the cells of the CAM are arranged so that each row of cells holds a memory address and that row of cells is connected by a match line to a corresponding word line of the data array to enable access of the data array in that word line when a match occurs on that match line.
- each row of the CAM holds the full address of a corresponding main memory location and the inputs to the CAM require the full address to be input.
- a prior art publication entitled “A Reconfigurable Content Addressable Memory,” by Steven A Guccione et al., discusses the implementation of a CAM within an FPGA.
- a CAM circuit 101 is very similar to a standard SRAM 100 .
- Both CAM and SRAM circuits are almost identical, each having 6 transistors 102 except for the addition of three match transistors 103 that provide for the parallel search capability of the CAM 101 .
- using standard programmable logic devices does not facilitate implementing such transistor level circuits.
- the prior art publication also addresses implementing of a CAM using look up tables (LUTs) in an FPGA. Rather than using flip-flops within the FPGA to store the data to be matched, this implementation addresses the use of LUTs for storing of the data to be matched. By using LUTs rather than flip-flops a smaller CAM architecture is possible.
- LUTs look up tables
- FIG. 2 illustrates the translation process from a virtual address (VA) 201 to a physical address (PA) 202 .
- VA 201 is a 32-bit address, VA[31:0]
- PA 202 is also a 32-bit address PA[31:0].
- the VA has two portions, a virtual page number (VPN) 203 and a page offset (PO) 204 .
- the VPN 203 is typically located in the upper portion of the VA 201 and the PA 202 is typically located in the lower portion, though this need not be so.
- the VPN is 20 bits and the PA is 12 bits.
- the PA, or lower 12 bits translate directly into the PA.
- the VPN 203 is used for indexing the TLB 205 to retrieve a physical page number (PPN) 206 therefrom. In other words, the VPN 203 undergoes translation to the PPN 206 . Combining the PPN 206 in the upper portion of the PA 202 and the PO into the lower portion of the PA provides a translation from the VA to the PA.
- PPN physical page number
- FIG. 3 illustrates the translation from a VA 201 to a PA 202 when performed in conjunction with a direct mapped cache memory 301 .
- the VA is used to access both the cache memory 301 and the TLB 205 .
- the page-offset portion of the VA is used to access the cache memory 301 —the page offset being the portion of the address that remains unmodified by the translation process.
- the page offset is used to index a tag array 302 and a data array 303 found in cache memory 301 where the page offset is used to index a cache line 302 a within the cache memory 301 .
- Access to the TLB 205 is performed using the VPN 203 portion of the VA 201 .
- the TLB 205 typically comprises a TLB tag array 304 and a TLB data array 305 .
- Both the TLB tag array 304 and the TLB data array 305 contain bits from the VPN 203 such that when a VPN is provided to both of these arrays, the bits making up the VPN are compared to those stored within the arrays 304 , 305 to locate an entry within the TLB 205 .
- the PPN 206 is retrieved and is provided to the cache memory 301 and used for comparison to the tag retrieved 302 a from the tag array 302 .
- a match being indicative of a cache “hit” 306 .
- a TLB hit signal 307 is generated. In this manner, the cache is only accessed using bits of the PPN 206 .
- the above example illustrates the use of a direct mapped cache memory; however, the same translation of a VA to a PA is applicable to set-associative caches as well. When set-associative caches are used, those of skill in the art appreciate that the size of a cache way is less than or equal to the size of a virtual page.
- FIG. 4 a generally illustrates a TLB 400 formed using synthesisable logic components 499 and a random access memory (RAM) 410 .
- a VPN for translation is provided via a VPN_IN input port 450 , where bits VPN_IN[31:12] are provided from the VA[31:0] to this input port 450 .
- a page mask signal is provided via a CPO_PAGE_MASK input port 451 .
- a CP0_TRANSLATION input signal is provided via a CP0_TRANSLATION input port 452 .
- a TLB_TRANSLATION output signal is provided via TLB_TRANSLATION output port 453 , in dependence upon a translation from a VA to a PA using the TLB 400 .
- FIG. 4 b illustrates a TLB 400 in more detail formed from synthesizeable logic components, and in FIG. 4 c the steps of operation for the TLB 400 are shown in summary.
- a VPN for translation is provided 480 via a VPN_IN input port 450 , where bits VPN_IN[31:12] are provided from the VA[31:0] to this input port 450 as the VPN.
- a page mask is provided via a CP0_PAGE_MASK input port 451 .
- This page mask is provided to a page mask encoder 408 , for encoding the page mask according to Table 1.
- TABLE 1 Page Mask Encoding page size mask[2:0] 4 kB 000 16 kB 001 64 kB 010 256 kB 011 1 M 100 4 M 101 16 M 110
- the page mask encoder 408 is used for accepting the CPO_PAGE_MASK input signal on an input port thereof and for correlating this input signal to a 3-bit vector, MASK[2:0].
- the 3-bit vector MASK[2:0] is further provided to a hashing circuit 406 .
- the hashing circuit 406 receives VPN_IN[31:12] via a first input port 406 a and MASK[2:0] via a second input port 406 b .
- a hashed vector H_VPN[5:0] is provided from an output port 406 c thereof via a hashing operation 481 of the hashing circuit 406 .
- the hashed vector H_VPN[5:0] and the MASK[2:0] are further provided to each one of 48 registers 409 , where each register consists of multiple flip-flops collectively referred to as 491 .
- Each of the registers 409 has two output ports. A first output signal from a first output port thereof is provided to a comparator circuit 403 . A second output signal from a second output port is provided to the second input port 406 b on one of 48 hashing circuits 406 . The first input port on this hashing circuit receives VPN_IN[31:12].
- the hashing circuit 406 output port is coupled to one of 48 comparator circuits 403 for performing a comparison between the register output and the hashing circuit output signal.
- Each of the comparators in dependence upon a comparison of two input signals, provides a ‘1’ if the signals are the same and a ‘0’ if they are different.
- Output signals hit, from each of the 48 comparators is provided to one of 48 single bit 2-input multiplexers 411 .
- Outputs ports from each of the multiplexers are coupled to a flip-flop 404 .
- Each of the flip-flop 404 generates an output signal provided at the output ports labeled try 1 , where collectively these output signals try[0 . . . 47], for 0 ⁇ i ⁇ 47 are provided to a priority encoder circuit 401 .
- the priority encoder circuit is further coupled to a binary decoder circuit 402 , where the priority encoder circuit asserts a TLB_ENTRY[5:0] signal to the binary decoder circuit 402 and to the RAM 410 .
- Three output ports are provided within the TLB 400 , an ENTRY_FOUND output port 454 , an ENTRY_NOT_FOUND output port 455 and a TLB_TRANSLATION output port 453 , for providing ENTRY_FOUND, ENTRY_NOT_FOUND, and TLB_TRANSLATION output signal, respectively.
- An address for translation from a VA to a PA is stored in a random access memory (RAM) 410 , with the RAM 410 preferably having 48-entries, in the form of lines.
- RAM random access memory
- input signals VPN_IN, CPO_PAGE_MASK, and CP0_TRANSLATION are provided to the TLB circuit 400 via input ports 450 , 451 , and 452 , respectively.
- Translations performed by the TLB are stored in RAM 410 for a given index, i. The given index, indexes one of the lines 410 a within the RAM that holds the translation to the PPN.
- the hashing circuit 406 computes the hash function H (VPN_IN, MASK) and stores the result in a corresponding 6-bit register h 1 490 .
- the page mask is stored in the 3-bit register m i 491 .
- a VPN is provided via the input port 450 and the hash functions H (VPN_IN, m,) is computed for all i and compared to h 1 . This yields a 48 bit vector 492 hit 0 . . . hit 47 which is subsequently loaded into a 48 bit register 493 try 0 . . . try 47 .
- H hash functions
- the priority encoder 401 selects the entry with the lowest index to address entries within the RAM.
- the decoder 402 converts this index to a 48-bit one-hot vector 494 clr 0 . . . clr 47 .
- the try, vector is reloaded, except for a bit corresponding to an index just used to address the RAM, which is cleared. This process is repeated, one entry at a time 483 . The process stops as soon as the requested entry is found 484 , as indicated by the ENTRY_FOUND signal on the ENTRY_FOUND output port 454 , or when all bits in try, are 0.
- the ENTRY_NOT_FOUND signal is provided via the ENTRY_NOT_FOUND output port 455 .
- the translation is successful and information for the translation is provided 485 from the RAM 410 using a TLB_TRANSLATION signal on the TLB_TRANSLATION output port 453 .
- the translation is not successful and the TLB reports a TLB refill exception.
- FIG. 5 illustrates a hashing circuit 506 in more detail.
- a 7 to 1 multiplexer 501 uses the MASK[2:0] and VPN[31:12] input signals to the hashing circuit 506 .
- a 7 to 1 multiplexer 501 provides the H_VPN[5:0] output signal from the hashing circuit 506 in dependence upon the MASK[2:0] signal provided to the second input port 506 b .
- This hashing circuit selects the 6 least significant bits from the VPN. The selection is controlled by the page mask because the definition of “least significant” changes with the page size.
- the 6 least significant bits (LSB)s of the VPN are bits 22:17, but with a 16 kB page size the 6 LSBs are bits 19:14.
- the TLB 400 stores two adjacent virtual pages per TLB entry, called an odd/even pair, the 6 LSBs for a 4 kB page odd/even pair are bits 18:13.
- bit 12 decides whether to return the even (0) or odd (1) translation, and for a 16 kB odd/even pair the bits are 20:15.
- This hash function is redundant, since the ordering of bits H_VPN[5:0] is irrelevant.
- FIG. 6 exploits the fact that ordering of bits is irrelevant.
- FIG. 6 illustrates a variation of the hashing circuit shown in FIG. 5.
- a VPN[31:12] signal is provided to the first input port 606 a
- a MASK[2:0] signal is provided to the second input port 606 b .
- the mask signal MASK[2:0] is comprised of bits m0, m1, and m2.
- this hashing 606 circuit there are 3, 3 to 1 multiplexers 601 through 603 .
- the first multiplexer 601 receives the following bits, ⁇ m 2 , ⁇ overscore (m) ⁇ 2 (m 1 +m 0 ) ⁇ on its selection input ports, where bits from the VPN, VPN[13:14], VPN[19:20], VPN[25:26] are provided to multiplexer data input ports, 0 through 2, respectively. Multiplexer 601 thus provides bits 5 and 4 to the H_VPN[5:0] output signal.
- the second multiplexer 602 receives the following bits ⁇ m 2 (m 1 +m 0 ), ⁇ overscore (m) ⁇ 2 m 1 +m 2 ⁇ overscore (m 1 m 0 ) ⁇ ) ⁇ on its selection input ports, where bits from the VPN, VPN[15:16], VPN[21:22], VPN[27:28] are provided to multiplexer data input ports, labeled 0 through 2 , respectively.
- Multiplexer 602 thus provides bits 3 and 2 to the H_VPN[5:0] output signal.
- the third multiplexer 603 receives the following bits ⁇ m 2 m 1 , ⁇ overscore (m 2 ) ⁇ m 1 m 0 +m 2 ⁇ overscore (m 1 ) ⁇ ) ⁇ on its selection input ports, where bits from the VPN, VPN[17:18], VPN[23:24], VPN[29:30] are provided to multiplexer data input ports, labeled 0 through 2 , respectively.
- Multiplexer 603 thus provides bits 1 and 0 to the H_VPN[5:0] output signal.
- the hash function H_VPN[5:0] is uniformly distributed for MASK[2:0] and for VPN_IN[31:12] input signals.
- VA when VA is provided to the TLB it is propagated to the synthesized logic for each line and a result is provided indicated by at least an asserted bit within the try 1 vector of bits. Only those lines for which a result indicative of a match occured are then physically accessed to provide the PPN As such only a small fraction of the TLB lines are accessed for the translation process, thus resulting in a substantial performance improvement.
Abstract
Description
- This invention relates to the area of translation lookaside buffers and more specifically to translation lookaside buffer architectures for rapid design cycles.
- Modern microprocessor systems typically utilize virtual addressing. Virtual addressing enables the system to effectively create a virtual memory space larger than an actual physical memory space. The process of breaking up the actual physical memory space into the virtual memory space is termed paging. Paging breaks up a linear address space of the physical memory space into fixed blocks called pages. Pages allow a large linear address space to be implemented with a smaller physical main memory plus cheap background memory. This configuration is referred to as “virtual memory.” Paging allows virtual memory to be implemented by managing memory in pages that are swapped to and from the background memory. Paging offers additional advantages, including reduced main memory fragmentation, selective memory write policies for different pages, and varying memory protection schemes for different pages. The presence of a paging mechanism is typically transparent to the application program.
- The size of a page is a tradeoff between flexibility and performance. A small page size allows finer control over the virtual memory system but it increases the overhead from paging activity. Therefore many CPUs support a mix of page sizes, e.g. a particular MIPS implementation supports any mix of 4 kB, 16 kB, 64 kB, 256 kB, 1 MB, 4 MB and 16 MB pages.
- A processor is then able to advantageously operate in the virtual address space using virtual addresses. Frequently, however, these virtual addresses must be translated into physical addresses—actual memory locations. One way of accomplishing this translation of virtual addresses into physical addresses is a use of translation tables that are regularly accessed and stored in main memory. Translation tables are stored in main memory because they are typically large in size. Unfortunately, regularly accessing of translation tables stored in main memory tends to slow overall system performance.
- Modern microprocessor systems often use a translation lookaside buffer (TLB) to store or cache recently generated virtual to physical address translations in order to avoid the need to regularly access translation tables in main memory to accomplish address translation. A TLB is a special type of cache memory. As with other types of cache memories, a TLB is typically comprised of a relatively small amount of memory storage specially designed to be quickly accessible. A TLB typically incorporates both a tag array and a data array, as are provided in cache memories. Within the tag array, each tag line stores a virtual address. This tag line is then associated with a corresponding data line in the data array in which is stored a physical address translation for the virtual address. Thus, prior to seeking a translation of a virtual address from translation tables in main memory, a processor first refers to the TLB to determine whether the physical address translation of the virtual address is presently stored in the TLB. In the event that the virtual address and corresponding physical address are stored in the TLB, the TLB provides the corresponding physical address at an output port thereof, and a time and resource-consuming access of main memory is avoided. To facilitate operation of the TLB and to reduce indexing requirements therefore, a content addressable memory (CAM) is typically provided within the TLB. CAMs are parallel pattern matching circuits. In a matching mode of operation the CAM permits searching of all of its data in parallel to find a match.
- Unfortunately, traditional TLBs require custom circuit design techniques to implement a CAM. Using custom circuit designs is not advantageous since each TLB and associated CAM requires a significant design effort in order to implement same in a processor system design. Of course, when a processor is absent CAM circuitry, signals from the processor propagate off chip to the CAM, thereby incurring delays.
- It is therefore an object of this invention to provide a CAM architecture formed of traditional synthesisable circuit blocks.
- In accordance with the invention there is provided a translation lookaside buffer (TLB) comprising: at least an input port for receiving a portion of a virtual address;
- a random access memory; a set of registers; and, synthesisable logic for determining a hash value from the received portion of the virtual address and for comparing the hash value to a stored hash value within the set of registers to determine a potential that a physical address associated with the virtual address is stored within a line within the random access memory and associated with a register, from the set of registers, within which the hash value is stored.
- In accordance with an aspect of the invention there is provided a translation lookaside buffer comprising: a random access memory; a first register associated with a line in the memory; and, a hashing circuit for receiving a virtual address other than a virtual address for which a translation is presently stored in the memory, for determining a hash value and for storing the hash value in the first register; and the hashing circuit for storing the virtual address and a translation therefor in the line in memory.
- In accordance with yet another aspect of the invention there is provided a translation lookaside buffer comprising: RAM; and, synthesisable logic for determining from a virtual address at least one potential address within the RAM in fixed relation to which to search for a physical address associated with the virtual address, the at least one potential address being other than the one and only known address within the RAM in fixed relation to which the physical address associated with the virtual address is stored.
- In accordance with yet another aspect of the invention there is provided a method of performing a virtual address lookup function for a translation lookaside buffer including RAM and synthesisable logic including the steps of: providing a virtual address to the synthesisable logic; hashing the provided virtual address to provide a hash result;
- based on the hash result determining a memory location within the RAM relative to which is stored a virtual address identifier and a physical address related thereto;
- comparing the virtual address to the virtual address identifier to determine if the physical address corresponds to the provided virtual address; and, when the physical address corresponds to the provided virtual address, providing the physical address as an output value.
- The invention will now be described with reference to the drawings in which:
- FIG. 1a illustrates a prior art transistor implementation of a SRAM circuit;
- FIG. 1b illustrates a prior art transistor implementation of a CAM circuit;
- FIG. 2 illustrates a prior art translation process from a virtual address (VA) to a physical address (PA);
- FIG. 3 illustrates a prior art translation from a VA to a PA when performed in conjunction with a direct mapped cache memory;
- FIG. 4a generally illustrates a translation lookaside buffer formed using synthesisable logic components and a random access memory;
- FIG. 4b illustrates a translation lookaside buffer in more detail formed from synthesisable logic components;
- FIG. 4c outlines the steps taken for operation of the TLB;
- FIG. 5 illustrates a hashing circuit in more detail; and,
- FIG. 6 illustrates a variation of the hashing circuit shown in FIG. 5.
- CAM circuits include storage circuits similar in structure to SRAM circuits. However, CAM circuits also include search circuitry offering an added benefit of a parallel search mode of operation, thus enabling searching of the contents of the CAM in parallel using hardware. When searching the CAM for a particular data value, the CAM provides a match signal upon finding a match for that data value within the CAM. A main difference between CAM and SRAM is that in a CAM, data is presented to the CAM representative of a virtual address and an address relating to the data is returned, whereas in a SRAM, an address is provided to the SRAM and data stored at that address is returned.
- The cells of the CAM are arranged so that each row of cells holds a memory address and that row of cells is connected by a match line to a corresponding word line of the data array to enable access of the data array in that word line when a match occurs on that match line. In a fully associative cache each row of the CAM holds the full address of a corresponding main memory location and the inputs to the CAM require the full address to be input.
- A prior art publication, entitled “A Reconfigurable Content Addressable Memory,” by Steven A Guccione et al., discusses the implementation of a CAM within an FPGA. As is seen in Prior Art FIG. 1, at a transistor level, the implementation of a
CAM circuit 101 is very similar to astandard SRAM 100. Both CAM and SRAM circuits are almost identical, each having 6transistors 102 except for the addition of threematch transistors 103 that provide for the parallel search capability of theCAM 101. Unfortunately, using standard programmable logic devices does not facilitate implementing such transistor level circuits. - In the prior art publication the implementation of the CAM in an FPGA is discussed. Using gate level logic to implement a CAM often results in an undesirable size of the CAM. Flip-flops are used as the data storage elements within the CAM and as a result the size of the CAM circuit attainable using an FPGA is dependent upon the number of flip-flops available within the FPGA. Implementing the CAM in an FPGA quickly depletes many of the FPGA resources and as a result is not a viable solution. Unfortunately this has lead prior designers to conclude that the CAM is only efficiently implemented at a transistor level.
- The prior art publication also addresses implementing of a CAM using look up tables (LUTs) in an FPGA. Rather than using flip-flops within the FPGA to store the data to be matched, this implementation addresses the use of LUTs for storing of the data to be matched. By using LUTs rather than flip-flops a smaller CAM architecture is possible.
- Unfortunately, forming CAMs from synthesisable elements is not easily done so prior art processors that offer CAM are provided with a CAM core within the processor. Providing a CAM core within the processor unfortunately makes the resulting circuit expensive because of the added design complexity. Such additional design complexity is ill-suited for small batch custom design processors.
- FIG. 2 illustrates the translation process from a virtual address (VA)201 to a physical address (PA) 202. The
VA 201 is a 32-bit address, VA[31:0], and thePA 202 is also a 32-bit address PA[31:0]. The VA has two portions, a virtual page number (VPN) 203 and a page offset (PO) 204. TheVPN 203 is typically located in the upper portion of theVA 201 and thePA 202 is typically located in the lower portion, though this need not be so. Typically for a 32-bit addressing scheme, the VPN is 20 bits and the PA is 12 bits. The PA, or lower 12 bits translate directly into the PA. TheVPN 203 is used for indexing theTLB 205 to retrieve a physical page number (PPN) 206 therefrom. In other words, theVPN 203 undergoes translation to thePPN 206. Combining thePPN 206 in the upper portion of thePA 202 and the PO into the lower portion of the PA provides a translation from the VA to the PA. - FIG. 3 illustrates the translation from a
VA 201 to aPA 202 when performed in conjunction with a direct mappedcache memory 301. At the beginning of a translation cycle, the VA is used to access both thecache memory 301 and theTLB 205. The page-offset portion of the VA is used to access thecache memory 301—the page offset being the portion of the address that remains unmodified by the translation process. The page offset is used to index atag array 302 and adata array 303 found incache memory 301 where the page offset is used to index acache line 302 a within thecache memory 301. Access to theTLB 205 is performed using theVPN 203 portion of theVA 201. TheTLB 205 typically comprises aTLB tag array 304 and aTLB data array 305. Both theTLB tag array 304 and theTLB data array 305 contain bits from theVPN 203 such that when a VPN is provided to both of these arrays, the bits making up the VPN are compared to those stored within thearrays TLB 205. - Once the
TLB data array 305 is accessed and a match is found between the VPN and an entry within theTLB data array 305 a, thePPN 206 is retrieved and is provided to thecache memory 301 and used for comparison to the tag retrieved 302 a from thetag array 302. A match being indicative of a cache “hit” 306. If a match is found between theVPN 203 and an entry within theTLB tag array 304 a then a TLB hitsignal 307 is generated. In this manner, the cache is only accessed using bits of thePPN 206. The above example illustrates the use of a direct mapped cache memory; however, the same translation of a VA to a PA is applicable to set-associative caches as well. When set-associative caches are used, those of skill in the art appreciate that the size of a cache way is less than or equal to the size of a virtual page. - Unfortunately, when a TLB is implemented in SRAM, an exhaustive search of the memory is required to support CAM functionality. Thus, when a TLB has storage for 1024 virtual addresses and their corresponding Physical Address, each address translation requires up to 1024 memory access and comparison operations. Such a CAM implementation is unworkable as the performance drops linearly with CAM size.
- FIG. 4a generally illustrates a
TLB 400 formed usingsynthesisable logic components 499 and a random access memory (RAM) 410. A VPN for translation is provided via aVPN_IN input port 450, where bits VPN_IN[31:12] are provided from the VA[31:0] to thisinput port 450. A page mask signal is provided via aCPO_PAGE_MASK input port 451. A CP0_TRANSLATION input signal is provided via aCP0_TRANSLATION input port 452. A TLB_TRANSLATION output signal is provided viaTLB_TRANSLATION output port 453, in dependence upon a translation from a VA to a PA using theTLB 400. - FIG. 4b illustrates a
TLB 400 in more detail formed from synthesizeable logic components, and in FIG. 4c the steps of operation for theTLB 400 are shown in summary. In more detailed description of the TLB operation, a VPN for translation is provided 480 via aVPN_IN input port 450, where bits VPN_IN[31:12] are provided from the VA[31:0] to thisinput port 450 as the VPN. A page mask is provided via aCP0_PAGE_MASK input port 451. This page mask is provided to apage mask encoder 408, for encoding the page mask according to Table 1.TABLE 1 Page Mask Encoding page size mask[2:0] 4 kB 000 16 kB 001 64 kB 010 256 kB 011 1 M 100 4 M 101 16 M 110 - The
page mask encoder 408 is used for accepting the CPO_PAGE_MASK input signal on an input port thereof and for correlating this input signal to a 3-bit vector, MASK[2:0]. The 3-bit vector MASK[2:0] is further provided to ahashing circuit 406. Thehashing circuit 406 receives VPN_IN[31:12] via afirst input port 406 a and MASK[2:0] via asecond input port 406 b. A hashed vector H_VPN[5:0] is provided from anoutput port 406 c thereof via ahashing operation 481 of thehashing circuit 406. The hashed vector H_VPN[5:0] and the MASK[2:0] are further provided to each one of 48registers 409, where each register consists of multiple flip-flops collectively referred to as 491. Each of theregisters 409 has two output ports. A first output signal from a first output port thereof is provided to acomparator circuit 403. A second output signal from a second output port is provided to thesecond input port 406b on one of 48 hashingcircuits 406. The first input port on this hashing circuit receives VPN_IN[31:12]. Thehashing circuit 406 output port is coupled to one of 48comparator circuits 403 for performing a comparison between the register output and the hashing circuit output signal. Each of the comparators, in dependence upon a comparison of two input signals, provides a ‘1’ if the signals are the same and a ‘0’ if they are different. Output signals hit, from each of the 48 comparators is provided to one of 48 single bit 2-input multiplexers 411. Outputs ports from each of the multiplexers are coupled to a flip-flop 404. Each of the flip-flop 404 generates an output signal provided at the output ports labeled try1, where collectively these output signals try[0 . . . 47], for 0≦i≦47 are provided to apriority encoder circuit 401. The priority encoder circuit is further coupled to abinary decoder circuit 402, where the priority encoder circuit asserts a TLB_ENTRY[5:0] signal to thebinary decoder circuit 402 and to theRAM 410. Three output ports are provided within theTLB 400, anENTRY_FOUND output port 454, anENTRY_NOT_FOUND output port 455 and aTLB_TRANSLATION output port 453, for providing ENTRY_FOUND, ENTRY_NOT_FOUND, and TLB_TRANSLATION output signal, respectively. - An address for translation from a VA to a PA is stored in a random access memory (RAM)410, with the
RAM 410 preferably having 48-entries, in the form of lines. In use, whenever a new translation is to be performed, input signals VPN_IN, CPO_PAGE_MASK, and CP0_TRANSLATION are provided to theTLB circuit 400 viainput ports RAM 410 for a given index, i. The given index, indexes one of thelines 410 a within the RAM that holds the translation to the PPN. Thehashing circuit 406 computes the hash function H (VPN_IN, MASK) and stores the result in a corresponding 6-bit register h 1 490. The page mask is stored in the 3-bit register m i 491. - When a translation is requested using the TLB, a VPN is provided via the
input port 450 and the hash functions H (VPN_IN, m,) is computed for all i and compared to h1. This yields a 48bit vector 492 hit0. . . hit47 which is subsequently loaded into a 48 bit register 493 try0. . . try47. In order to determine whether the requested VPN_IN is present in the translation table stored inRAM 482, only those entries, or lines, in RAM are checked for which tryi is asserted. An entry in the 48-bit try, vector is asserted if it yields a ‘1’ 483. Of course, there may be more than one bit asserted in the try, vector, but thepriority encoder 401 selects the entry with the lowest index to address entries within the RAM. Thedecoder 402 converts this index to a 48-bit one-hot vector 494 clr0. . . clr47. When the clock pulse arrives from a clock circuit (not shown), the try, vector is reloaded, except for a bit corresponding to an index just used to address the RAM, which is cleared. This process is repeated, one entry at atime 483. The process stops as soon as the requested entry is found 484, as indicated by the ENTRY_FOUND signal on theENTRY_FOUND output port 454, or when all bits in try, are 0. When all bits in tryi are ‘0’ then the ENTRY_NOT_FOUND signal is provided via theENTRY_NOT_FOUND output port 455. In the first case the translation is successful and information for the translation is provided 485 from theRAM 410 using a TLB_TRANSLATION signal on theTLB_TRANSLATION output port 453. In the second case the translation is not successful and the TLB reports a TLB refill exception. - FIG. 5 illustrates a
hashing circuit 506 in more detail. Using the MASK[2:0] and VPN[31:12] input signals to thehashing circuit 506, a 7 to 1multiplexer 501 provides the H_VPN[5:0] output signal from the hashingcircuit 506 in dependence upon the MASK[2:0] signal provided to thesecond input port 506 b. This hashing circuit selects the 6 least significant bits from the VPN. The selection is controlled by the page mask because the definition of “least significant” changes with the page size. For example, with a 4 kB page size, the 6 least significant bits (LSB)s of the VPN are bits 22:17, but with a 16 kB page size the 6 LSBs are bits 19:14. Since theTLB 400 stores two adjacent virtual pages per TLB entry, called an odd/even pair, the 6 LSBs for a 4 kB page odd/even pair are bits 18:13. Thus bit 12 decides whether to return the even (0) or odd (1) translation, and for a 16 kB odd/even pair the bits are 20:15. This hash function, however, is redundant, since the ordering of bits H_VPN[5:0] is irrelevant. FIG. 6 exploits the fact that ordering of bits is irrelevant. - FIG. 6 illustrates a variation of the hashing circuit shown in FIG. 5. A VPN[31:12] signal is provided to the
first input port 606 a, and a MASK[2:0] signal is provided to thesecond input port 606 b. The mask signal MASK[2:0] is comprised of bits m0, m1, and m2. Within this hashing 606 circuit there are 3, 3 to 1multiplexers 601 through 603. Thefirst multiplexer 601 receives the following bits, {m2,{overscore (m)}2(m1+m0)} on its selection input ports, where bits from the VPN, VPN[13:14], VPN[19:20], VPN[25:26] are provided to multiplexer data input ports, 0 through 2, respectively.Multiplexer 601 thus providesbits second multiplexer 602 receives the following bits {m2(m1+m0),{overscore (m)}2m1+m2{overscore (m1m0)})} on its selection input ports, where bits from the VPN, VPN[15:16], VPN[21:22], VPN[27:28] are provided to multiplexer data input ports, labeled 0 through 2, respectively.Multiplexer 602 thus providesbits third multiplexer 603 receives the following bits {m2m1, {overscore (m2)}m1m0+m2{overscore (m1)})} on its selection input ports, where bits from the VPN, VPN[17:18], VPN[23:24], VPN[29:30] are provided to multiplexer data input ports, labeled 0 through 2, respectively.Multiplexer 603 thus providesbits -
-
- , gives the probability that exactly j bits in the try vector tryi are initially asserted. Having a uniform hashing function H with n bits at the output signal thereof, p=2−n, wherein the case of FIG. 4b, n=6.
- In the case of a TLB hit, at least one access to the
RAM 410 us required, as opposed to a TLB miss condition which is detected without accessing the RAM, since in a TLB miss condition the try vector try, contains all zeros. -
- For a TLB hit, there must be at least one ‘1’ in the try vector try1. The only uncertainty is with the remaining elements within the vector. The variable k is used to represent the number of remaining entries that are set to ‘1’ within the try vector try1 for k in the range from 0 . . . 47. If k=0 then only one entry within the RAM is looked up. Therefore, since one clock cycle was used to find the translation in the first location for i=0, then a total of two clock cycles are utilized to perform the translation. On average, it takes 2+k/2 cycles to return the requested translation from
RAM 410. -
- , gives the probability that the TLB will use 25 or more cycles to complete a translation. Table 2 lists, for a range of hash function widths (n), the average number of cycles it takes to find a translation Nhlt, to detect a miss Nmiss and the probability that the TLB operation takes 25 cycles or more.
TABLE 2 TLB latency as a function of the number of hash bits ‘n’ n Nhu Nhitq Nmiss P{N25} 3 4.94 3.94 7.00 4.3 10−11 4 3.46 2.46 4.00 5.9 10−18 5 2.73 1.73 2.50 3.6 10−25 6 2.37 1.37 1.75 1.5 10−32 7 2.18 1.18 1.38 5.4 10−40 - From Table 2 it is evident that P {N25} is so small that even with a 4 bit hash function it takes more than 6000 years of continuous operation to run into a case where the TLB translation requires between 25 and 48 clock cycles.
- The column Nhltq (“hit quick”) applies to the case where the VPN_IN is applied continuously to the
TLB circuit 400. From this table it is evident that having n=5 or n=6 is sufficient when focusing on the most important number, which is Nhlt. There is not much to be gained beyond 6 bits, since Nhit approaches 2.0 when n=>20. A value of n=6 is used in theTLB circuit 400 since the hash function may not be very uniform. Therefore, 6-bit hash function used within the TLB approximates the performance of a 5-bit truly uniform hash function. - Advantageously, when VA is provided to the TLB it is propagated to the synthesized logic for each line and a result is provided indicated by at least an asserted bit within the try1 vector of bits. Only those lines for which a result indicative of a match occured are then physically accessed to provide the PPN As such only a small fraction of the TLB lines are accessed for the translation process, thus resulting in a substantial performance improvement.
- Numerous other embodiments may be envisaged without departing from the spirit or scope of the invention.
Claims (34)
Priority Applications (8)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/242,785 US20040054867A1 (en) | 2002-09-13 | 2002-09-13 | Translation lookaside buffer |
KR1020057004205A KR101145557B1 (en) | 2002-09-13 | 2003-09-12 | Translation lookaside buffer with partial tag hashed according to page size indicator |
CNB038215691A CN100555248C (en) | 2002-09-13 | 2003-09-12 | Have the translation look aside buffer that carries out the local flag of hash according to the page size designator |
AU2003260822A AU2003260822A1 (en) | 2002-09-13 | 2003-09-12 | Translation lookaside buffer with partial tag hashed according to page size indicator |
JP2004535773A JP2005538465A (en) | 2002-09-13 | 2003-09-12 | Conversion index buffer with partial tags hashed based on page size index |
EP03795156A EP1552397A1 (en) | 2002-09-13 | 2003-09-12 | Translation lookaside buffer with partial tag hashed according to page size indicator |
PCT/IB2003/003915 WO2004025480A1 (en) | 2002-09-13 | 2003-09-12 | Translation lookaside buffer with partial tag hashed according to page size indicator |
US13/298,800 US8607026B2 (en) | 2002-09-13 | 2011-11-17 | Translation lookaside buffer |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/242,785 US20040054867A1 (en) | 2002-09-13 | 2002-09-13 | Translation lookaside buffer |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/298,800 Continuation US8607026B2 (en) | 2002-09-13 | 2011-11-17 | Translation lookaside buffer |
Publications (1)
Publication Number | Publication Date |
---|---|
US20040054867A1 true US20040054867A1 (en) | 2004-03-18 |
Family
ID=31991480
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/242,785 Abandoned US20040054867A1 (en) | 2002-09-13 | 2002-09-13 | Translation lookaside buffer |
US13/298,800 Expired - Lifetime US8607026B2 (en) | 2002-09-13 | 2011-11-17 | Translation lookaside buffer |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/298,800 Expired - Lifetime US8607026B2 (en) | 2002-09-13 | 2011-11-17 | Translation lookaside buffer |
Country Status (7)
Country | Link |
---|---|
US (2) | US20040054867A1 (en) |
EP (1) | EP1552397A1 (en) |
JP (1) | JP2005538465A (en) |
KR (1) | KR101145557B1 (en) |
CN (1) | CN100555248C (en) |
AU (1) | AU2003260822A1 (en) |
WO (1) | WO2004025480A1 (en) |
Cited By (34)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040078545A1 (en) * | 2002-10-22 | 2004-04-22 | Min-Su Kim | High-speed translation lookaside buffer |
US20050175018A1 (en) * | 2003-05-15 | 2005-08-11 | Wong Yuen F. | System and method for high speed packet transmission implementing dual transmit and receive pipelines |
US20070288690A1 (en) * | 2006-06-13 | 2007-12-13 | Foundry Networks, Inc. | High bandwidth, high capacity look-up table implementation in dynamic random access memory |
US20080002707A1 (en) * | 2002-05-06 | 2008-01-03 | Davis Ian E | Flexible method for processing data packets in a network routing system for enhanced efficiency and monitoring capability |
US20080049742A1 (en) * | 2006-08-22 | 2008-02-28 | Deepak Bansal | System and method for ecmp load sharing |
US20080140724A1 (en) * | 2006-12-06 | 2008-06-12 | David Flynn | Apparatus, system, and method for servicing object requests within a storage controller |
US20090134399A1 (en) * | 2005-02-18 | 2009-05-28 | Semiconductor Energy Laboratory Co., Ltd. | Semiconductor Device and Method for Manufacturing the Same |
US20090150605A1 (en) * | 2007-12-06 | 2009-06-11 | David Flynn | Apparatus, system, and method for converting a storage request into an append data storage command |
US20090279559A1 (en) * | 2004-03-26 | 2009-11-12 | Foundry Networks, Inc., A Delaware Corporation | Method and apparatus for aggregating input data streams |
US20090279441A1 (en) * | 2007-01-11 | 2009-11-12 | Foundry Networks, Inc. | Techniques for transmitting failure detection protocol packets |
US20090279561A1 (en) * | 2000-11-17 | 2009-11-12 | Foundry Networks, Inc. | Backplane Interface Adapter |
US20090282322A1 (en) * | 2007-07-18 | 2009-11-12 | Foundry Networks, Inc. | Techniques for segmented crc design in high speed networks |
US20090279548A1 (en) * | 2002-05-06 | 2009-11-12 | Foundry Networks, Inc. | Pipeline method and system for switching packets |
US20090279558A1 (en) * | 2002-05-06 | 2009-11-12 | Ian Edward Davis | Network routing apparatus for enhanced efficiency and monitoring capability |
US20090287952A1 (en) * | 2000-11-17 | 2009-11-19 | Foundry Networks, Inc. | Backplane Interface Adapter with Error Control and Redundant Fabric |
US7649885B1 (en) | 2002-05-06 | 2010-01-19 | Foundry Networks, Inc. | Network routing system for enhanced efficiency and monitoring capability |
US7657703B1 (en) | 2004-10-29 | 2010-02-02 | Foundry Networks, Inc. | Double density content addressable memory (CAM) lookup scheme |
US7738450B1 (en) | 2002-05-06 | 2010-06-15 | Foundry Networks, Inc. | System architecture for very fast ethernet blade |
US20110216769A1 (en) * | 2010-03-08 | 2011-09-08 | Brocade Communications Systems, Inc. | Dynamic Path Selection |
US8090901B2 (en) | 2009-05-14 | 2012-01-03 | Brocade Communications Systems, Inc. | TCAM management approach that minimize movements |
US8149839B1 (en) | 2007-09-26 | 2012-04-03 | Foundry Networks, Llc | Selection of trunk ports and paths using rotation |
US8161353B2 (en) | 2007-12-06 | 2012-04-17 | Fusion-Io, Inc. | Apparatus, system, and method for validating that a correct data segment is read from a data storage device |
US8238255B2 (en) | 2006-11-22 | 2012-08-07 | Foundry Networks, Llc | Recovering from failures without impact on data traffic in a shared bus architecture |
US8271859B2 (en) | 2007-07-18 | 2012-09-18 | Foundry Networks Llc | Segmented CRC design in high speed networks |
WO2013040335A1 (en) * | 2011-09-14 | 2013-03-21 | Apple Inc. | Efficient non-volatile read cache for storage system |
US20130111151A1 (en) * | 2011-10-27 | 2013-05-02 | Oracle International Corporation | Software translation lookaside buffer for persistent pointer management |
US8448162B2 (en) | 2005-12-28 | 2013-05-21 | Foundry Networks, Llc | Hitless software upgrades |
US8599850B2 (en) | 2009-09-21 | 2013-12-03 | Brocade Communications Systems, Inc. | Provisioning single or multistage networks using ethernet service instances (ESIs) |
US8671219B2 (en) | 2002-05-06 | 2014-03-11 | Foundry Networks, Llc | Method and apparatus for efficiently processing data packets in a computer network |
US8730961B1 (en) * | 2004-04-26 | 2014-05-20 | Foundry Networks, Llc | System and method for optimizing router lookup |
US20140223090A1 (en) * | 2013-02-01 | 2014-08-07 | Apple Inc | Accessing control registers over a data bus |
US9959044B2 (en) * | 2016-05-03 | 2018-05-01 | Macronix International Co., Ltd. | Memory device including risky mapping table and controlling method thereof |
US10545877B2 (en) * | 2018-04-05 | 2020-01-28 | Arm Limited | Apparatus and method for accessing an address translation cache |
EP3719656A1 (en) * | 2019-03-29 | 2020-10-07 | Intel Corporation | Valid bits of a translation lookaside buffer (tlb) for checking multiple page sizes in one probe cycle and reconfigurable sub-tlbs |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8775740B2 (en) * | 2004-08-30 | 2014-07-08 | Texas Instruments Incorporated | System and method for high performance, power efficient store buffer forwarding |
CN100520710C (en) * | 2006-07-27 | 2009-07-29 | 中国科学院计算技术研究所 | TLBR inside exception processing method in complicated instruction system and processor |
CN102486751A (en) * | 2010-12-01 | 2012-06-06 | 安凯(广州)微电子技术有限公司 | Method for realizing virtual big page through small page NANDFLASH on micro memory system |
CN102662860B (en) * | 2012-03-15 | 2015-07-01 | 天津国芯科技有限公司 | Translation lookaside buffer (TLB) for process switching and address matching method therein |
CN104216833B (en) | 2013-05-29 | 2017-10-10 | 华为技术有限公司 | A kind of method and device for determining physical address |
US10007435B2 (en) | 2015-05-21 | 2018-06-26 | Micron Technology, Inc. | Translation lookaside buffer in memory |
Citations (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4170039A (en) * | 1978-07-17 | 1979-10-02 | International Business Machines Corporation | Virtual address translation speed up technique |
US4215402A (en) * | 1978-10-23 | 1980-07-29 | International Business Machines Corporation | Hash index table hash generator apparatus |
US4638426A (en) * | 1982-12-30 | 1987-01-20 | International Business Machines Corporation | Virtual memory address translation mechanism with controlled data persistence |
US4680700A (en) * | 1983-12-07 | 1987-07-14 | International Business Machines Corporation | Virtual memory address translation mechanism with combined hash address table and inverted page table |
US5136702A (en) * | 1989-05-23 | 1992-08-04 | Hitachi, Ltd. | Buffer storage control method and apparatus |
US5526504A (en) * | 1993-12-15 | 1996-06-11 | Silicon Graphics, Inc. | Variable page size translation lookaside buffer |
US5574875A (en) * | 1992-03-13 | 1996-11-12 | Inmos Limited | Cache memory system including a RAM for storing data and CAM cell arrays for storing virtual and physical addresses |
US5574877A (en) * | 1992-09-25 | 1996-11-12 | Silicon Graphics, Inc. | TLB with two physical pages per virtual tag |
US5752275A (en) * | 1995-03-31 | 1998-05-12 | Intel Corporation | Translation look-aside buffer including a single page size translation unit |
US5860147A (en) * | 1996-09-16 | 1999-01-12 | Intel Corporation | Method and apparatus for replacement of entries in a translation look-aside buffer |
US6014732A (en) * | 1997-10-22 | 2000-01-11 | Hewlett-Packard Company | Cache memory with reduced access time |
US6026476A (en) * | 1996-03-19 | 2000-02-15 | Intel Corporation | Fast fully associative translation lookaside buffer |
US6205531B1 (en) * | 1998-07-02 | 2001-03-20 | Silicon Graphics Incorporated | Method and apparatus for virtual address translation |
US6233652B1 (en) * | 1998-10-30 | 2001-05-15 | Intel Corporation | Translation lookaside buffer for multiple page sizes |
US6360220B1 (en) * | 1998-08-04 | 2002-03-19 | Microsoft Corporation | Lock-free methods and systems for accessing and storing information in an indexed computer data structure having modifiable entries |
US6381673B1 (en) * | 1998-07-06 | 2002-04-30 | Netlogic Microsystems, Inc. | Method and apparatus for performing a read next highest priority match instruction in a content addressable memory device |
US20020073073A1 (en) * | 2000-12-07 | 2002-06-13 | Cheng Paul C. | Paralleled content addressable memory search engine |
US20030037055A1 (en) * | 2001-08-09 | 2003-02-20 | Paul Cheng | Large database search using CAM and hash |
US6625715B1 (en) * | 1999-12-30 | 2003-09-23 | Intel Corporation | System and method for translation buffer accommodating multiple page sizes |
US20030233515A1 (en) * | 2002-06-14 | 2003-12-18 | Integrated Device Technology, Inc. | Hardware hashing of an input of a content addressable memory (CAM) to emulate a wider CAM |
US20040170039A1 (en) * | 2001-10-31 | 2004-09-02 | Mobility Electronics Inc. | Dual input AC and DC power supply having a programmable DC output utilizing a secondary buck converter |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0675443A1 (en) * | 1994-03-30 | 1995-10-04 | Digital Equipment Corporation | Apparatus and method for accessing direct mapped cache |
US6516386B1 (en) * | 1997-12-31 | 2003-02-04 | Intel Corporation | Method and apparatus for indexing a cache |
US6212603B1 (en) * | 1998-04-09 | 2001-04-03 | Institute For The Development Of Emerging Architectures, L.L.C. | Processor with apparatus for tracking prefetch and demand fetch instructions serviced by cache memory |
US6625714B1 (en) * | 1999-12-17 | 2003-09-23 | Hewlett-Packard Development Company, L.P. | Parallel distributed function translation lookaside buffer |
US6687789B1 (en) * | 2000-01-03 | 2004-02-03 | Advanced Micro Devices, Inc. | Cache which provides partial tags from non-predicted ways to direct search if way prediction misses |
US6356990B1 (en) * | 2000-02-02 | 2002-03-12 | International Business Machines Corporation | Set-associative cache memory having a built-in set prediction array |
US6581140B1 (en) * | 2000-07-03 | 2003-06-17 | Motorola, Inc. | Method and apparatus for improving access time in set-associative cache systems |
US6925464B2 (en) * | 2002-06-13 | 2005-08-02 | Intel Corporation | Method and system for performing inserts and lookups in memory |
-
2002
- 2002-09-13 US US10/242,785 patent/US20040054867A1/en not_active Abandoned
-
2003
- 2003-09-12 WO PCT/IB2003/003915 patent/WO2004025480A1/en active Application Filing
- 2003-09-12 KR KR1020057004205A patent/KR101145557B1/en active IP Right Grant
- 2003-09-12 EP EP03795156A patent/EP1552397A1/en not_active Withdrawn
- 2003-09-12 CN CNB038215691A patent/CN100555248C/en not_active Expired - Fee Related
- 2003-09-12 AU AU2003260822A patent/AU2003260822A1/en not_active Abandoned
- 2003-09-12 JP JP2004535773A patent/JP2005538465A/en active Pending
-
2011
- 2011-11-17 US US13/298,800 patent/US8607026B2/en not_active Expired - Lifetime
Patent Citations (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4170039A (en) * | 1978-07-17 | 1979-10-02 | International Business Machines Corporation | Virtual address translation speed up technique |
US4215402A (en) * | 1978-10-23 | 1980-07-29 | International Business Machines Corporation | Hash index table hash generator apparatus |
US4638426A (en) * | 1982-12-30 | 1987-01-20 | International Business Machines Corporation | Virtual memory address translation mechanism with controlled data persistence |
US4680700A (en) * | 1983-12-07 | 1987-07-14 | International Business Machines Corporation | Virtual memory address translation mechanism with combined hash address table and inverted page table |
US5136702A (en) * | 1989-05-23 | 1992-08-04 | Hitachi, Ltd. | Buffer storage control method and apparatus |
US5574875A (en) * | 1992-03-13 | 1996-11-12 | Inmos Limited | Cache memory system including a RAM for storing data and CAM cell arrays for storing virtual and physical addresses |
US5574877A (en) * | 1992-09-25 | 1996-11-12 | Silicon Graphics, Inc. | TLB with two physical pages per virtual tag |
US5526504A (en) * | 1993-12-15 | 1996-06-11 | Silicon Graphics, Inc. | Variable page size translation lookaside buffer |
US5752275A (en) * | 1995-03-31 | 1998-05-12 | Intel Corporation | Translation look-aside buffer including a single page size translation unit |
US6026476A (en) * | 1996-03-19 | 2000-02-15 | Intel Corporation | Fast fully associative translation lookaside buffer |
US5860147A (en) * | 1996-09-16 | 1999-01-12 | Intel Corporation | Method and apparatus for replacement of entries in a translation look-aside buffer |
US6014732A (en) * | 1997-10-22 | 2000-01-11 | Hewlett-Packard Company | Cache memory with reduced access time |
US6205531B1 (en) * | 1998-07-02 | 2001-03-20 | Silicon Graphics Incorporated | Method and apparatus for virtual address translation |
US6381673B1 (en) * | 1998-07-06 | 2002-04-30 | Netlogic Microsystems, Inc. | Method and apparatus for performing a read next highest priority match instruction in a content addressable memory device |
US6360220B1 (en) * | 1998-08-04 | 2002-03-19 | Microsoft Corporation | Lock-free methods and systems for accessing and storing information in an indexed computer data structure having modifiable entries |
US6233652B1 (en) * | 1998-10-30 | 2001-05-15 | Intel Corporation | Translation lookaside buffer for multiple page sizes |
US6625715B1 (en) * | 1999-12-30 | 2003-09-23 | Intel Corporation | System and method for translation buffer accommodating multiple page sizes |
US20020073073A1 (en) * | 2000-12-07 | 2002-06-13 | Cheng Paul C. | Paralleled content addressable memory search engine |
US20030037055A1 (en) * | 2001-08-09 | 2003-02-20 | Paul Cheng | Large database search using CAM and hash |
US20040170039A1 (en) * | 2001-10-31 | 2004-09-02 | Mobility Electronics Inc. | Dual input AC and DC power supply having a programmable DC output utilizing a secondary buck converter |
US20030233515A1 (en) * | 2002-06-14 | 2003-12-18 | Integrated Device Technology, Inc. | Hardware hashing of an input of a content addressable memory (CAM) to emulate a wider CAM |
Cited By (79)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090287952A1 (en) * | 2000-11-17 | 2009-11-19 | Foundry Networks, Inc. | Backplane Interface Adapter with Error Control and Redundant Fabric |
US20090279561A1 (en) * | 2000-11-17 | 2009-11-12 | Foundry Networks, Inc. | Backplane Interface Adapter |
US7948872B2 (en) | 2000-11-17 | 2011-05-24 | Foundry Networks, Llc | Backplane interface adapter with error control and redundant fabric |
US8514716B2 (en) | 2000-11-17 | 2013-08-20 | Foundry Networks, Llc | Backplane interface adapter with error control and redundant fabric |
US8619781B2 (en) | 2000-11-17 | 2013-12-31 | Foundry Networks, Llc | Backplane interface adapter with error control and redundant fabric |
US20100034215A1 (en) * | 2000-11-17 | 2010-02-11 | Foundry Networks, Inc. | Backplane Interface Adapter with Error Control |
US7978702B2 (en) | 2000-11-17 | 2011-07-12 | Foundry Networks, Llc | Backplane interface adapter |
US20090290499A1 (en) * | 2000-11-17 | 2009-11-26 | Foundry Networks, Inc. | Backplane Interface Adapter with Error Control and Redundant Fabric |
US8964754B2 (en) | 2000-11-17 | 2015-02-24 | Foundry Networks, Llc | Backplane interface adapter with error control and redundant fabric |
US9030937B2 (en) | 2000-11-17 | 2015-05-12 | Foundry Networks, Llc | Backplane interface adapter with error control and redundant fabric |
US7995580B2 (en) | 2000-11-17 | 2011-08-09 | Foundry Networks, Inc. | Backplane interface adapter with error control and redundant fabric |
US8194666B2 (en) | 2002-05-06 | 2012-06-05 | Foundry Networks, Llc | Flexible method for processing data packets in a network routing system for enhanced efficiency and monitoring capability |
US7738450B1 (en) | 2002-05-06 | 2010-06-15 | Foundry Networks, Inc. | System architecture for very fast ethernet blade |
US20090279548A1 (en) * | 2002-05-06 | 2009-11-12 | Foundry Networks, Inc. | Pipeline method and system for switching packets |
US20090279558A1 (en) * | 2002-05-06 | 2009-11-12 | Ian Edward Davis | Network routing apparatus for enhanced efficiency and monitoring capability |
US8170044B2 (en) | 2002-05-06 | 2012-05-01 | Foundry Networks, Llc | Pipeline method and system for switching packets |
US7830884B2 (en) | 2002-05-06 | 2010-11-09 | Foundry Networks, Llc | Flexible method for processing data packets in a network routing system for enhanced efficiency and monitoring capability |
US8989202B2 (en) | 2002-05-06 | 2015-03-24 | Foundry Networks, Llc | Pipeline method and system for switching packets |
US7649885B1 (en) | 2002-05-06 | 2010-01-19 | Foundry Networks, Inc. | Network routing system for enhanced efficiency and monitoring capability |
US8671219B2 (en) | 2002-05-06 | 2014-03-11 | Foundry Networks, Llc | Method and apparatus for efficiently processing data packets in a computer network |
US7813367B2 (en) | 2002-05-06 | 2010-10-12 | Foundry Networks, Inc. | Pipeline method and system for switching packets |
US20080002707A1 (en) * | 2002-05-06 | 2008-01-03 | Davis Ian E | Flexible method for processing data packets in a network routing system for enhanced efficiency and monitoring capability |
US20040078545A1 (en) * | 2002-10-22 | 2004-04-22 | Min-Su Kim | High-speed translation lookaside buffer |
US7035992B2 (en) * | 2002-10-22 | 2006-04-25 | Samsung Electronics Co., Ltd. | High-speed translation lookaside buffer |
US9461940B2 (en) | 2003-05-15 | 2016-10-04 | Foundry Networks, Llc | System and method for high speed packet transmission |
US8718051B2 (en) | 2003-05-15 | 2014-05-06 | Foundry Networks, Llc | System and method for high speed packet transmission |
US20050175018A1 (en) * | 2003-05-15 | 2005-08-11 | Wong Yuen F. | System and method for high speed packet transmission implementing dual transmit and receive pipelines |
US8811390B2 (en) | 2003-05-15 | 2014-08-19 | Foundry Networks, Llc | System and method for high speed packet transmission |
US9338100B2 (en) | 2004-03-26 | 2016-05-10 | Foundry Networks, Llc | Method and apparatus for aggregating input data streams |
US7817659B2 (en) | 2004-03-26 | 2010-10-19 | Foundry Networks, Llc | Method and apparatus for aggregating input data streams |
US20090279559A1 (en) * | 2004-03-26 | 2009-11-12 | Foundry Networks, Inc., A Delaware Corporation | Method and apparatus for aggregating input data streams |
US8493988B2 (en) | 2004-03-26 | 2013-07-23 | Foundry Networks, Llc | Method and apparatus for aggregating input data streams |
US8730961B1 (en) * | 2004-04-26 | 2014-05-20 | Foundry Networks, Llc | System and method for optimizing router lookup |
US7953922B2 (en) | 2004-10-29 | 2011-05-31 | Foundry Networks, Llc | Double density content addressable memory (CAM) lookup scheme |
US7657703B1 (en) | 2004-10-29 | 2010-02-02 | Foundry Networks, Inc. | Double density content addressable memory (CAM) lookup scheme |
US20100100671A1 (en) * | 2004-10-29 | 2010-04-22 | Foundry Networks, Inc. | Double density content addressable memory (cam) lookup scheme |
US7953923B2 (en) | 2004-10-29 | 2011-05-31 | Foundry Networks, Llc | Double density content addressable memory (CAM) lookup scheme |
US20090134399A1 (en) * | 2005-02-18 | 2009-05-28 | Semiconductor Energy Laboratory Co., Ltd. | Semiconductor Device and Method for Manufacturing the Same |
US9378005B2 (en) | 2005-12-28 | 2016-06-28 | Foundry Networks, Llc | Hitless software upgrades |
US8448162B2 (en) | 2005-12-28 | 2013-05-21 | Foundry Networks, Llc | Hitless software upgrades |
US20070288690A1 (en) * | 2006-06-13 | 2007-12-13 | Foundry Networks, Inc. | High bandwidth, high capacity look-up table implementation in dynamic random access memory |
US20080049742A1 (en) * | 2006-08-22 | 2008-02-28 | Deepak Bansal | System and method for ecmp load sharing |
US20110044340A1 (en) * | 2006-08-22 | 2011-02-24 | Foundry Networks, Llc | System and method for ecmp load sharing |
US7903654B2 (en) | 2006-08-22 | 2011-03-08 | Foundry Networks, Llc | System and method for ECMP load sharing |
US9030943B2 (en) | 2006-11-22 | 2015-05-12 | Foundry Networks, Llc | Recovering from failures without impact on data traffic in a shared bus architecture |
US8238255B2 (en) | 2006-11-22 | 2012-08-07 | Foundry Networks, Llc | Recovering from failures without impact on data traffic in a shared bus architecture |
US11960412B2 (en) | 2006-12-06 | 2024-04-16 | Unification Technologies Llc | Systems and methods for identifying storage resources that are not in use |
US11640359B2 (en) | 2006-12-06 | 2023-05-02 | Unification Technologies Llc | Systems and methods for identifying storage resources that are not in use |
US11847066B2 (en) | 2006-12-06 | 2023-12-19 | Unification Technologies Llc | Apparatus, system, and method for managing commands of solid-state storage using bank interleave |
US20080140724A1 (en) * | 2006-12-06 | 2008-06-12 | David Flynn | Apparatus, system, and method for servicing object requests within a storage controller |
US8392798B2 (en) | 2006-12-06 | 2013-03-05 | Fusion-Io, Inc. | Apparatus, system, and method for validating that correct data is read from a storage device |
US11573909B2 (en) | 2006-12-06 | 2023-02-07 | Unification Technologies Llc | Apparatus, system, and method for managing commands of solid-state storage using bank interleave |
US20090279441A1 (en) * | 2007-01-11 | 2009-11-12 | Foundry Networks, Inc. | Techniques for transmitting failure detection protocol packets |
US20090279542A1 (en) * | 2007-01-11 | 2009-11-12 | Foundry Networks, Inc. | Techniques for using dual memory structures for processing failure detection protocol packets |
US8395996B2 (en) | 2007-01-11 | 2013-03-12 | Foundry Networks, Llc | Techniques for processing incoming failure detection protocol packets |
US7978614B2 (en) | 2007-01-11 | 2011-07-12 | Foundry Network, LLC | Techniques for detecting non-receipt of fault detection protocol packets |
US9112780B2 (en) | 2007-01-11 | 2015-08-18 | Foundry Networks, Llc | Techniques for processing incoming failure detection protocol packets |
US8155011B2 (en) | 2007-01-11 | 2012-04-10 | Foundry Networks, Llc | Techniques for using dual memory structures for processing failure detection protocol packets |
US8271859B2 (en) | 2007-07-18 | 2012-09-18 | Foundry Networks Llc | Segmented CRC design in high speed networks |
US20090282322A1 (en) * | 2007-07-18 | 2009-11-12 | Foundry Networks, Inc. | Techniques for segmented crc design in high speed networks |
US8037399B2 (en) | 2007-07-18 | 2011-10-11 | Foundry Networks, Llc | Techniques for segmented CRC design in high speed networks |
US8149839B1 (en) | 2007-09-26 | 2012-04-03 | Foundry Networks, Llc | Selection of trunk ports and paths using rotation |
US8509236B2 (en) | 2007-09-26 | 2013-08-13 | Foundry Networks, Llc | Techniques for selecting paths and/or trunk ports for forwarding traffic flows |
US20090150605A1 (en) * | 2007-12-06 | 2009-06-11 | David Flynn | Apparatus, system, and method for converting a storage request into an append data storage command |
US8151082B2 (en) | 2007-12-06 | 2012-04-03 | Fusion-Io, Inc. | Apparatus, system, and method for converting a storage request into an append data storage command |
US8161353B2 (en) | 2007-12-06 | 2012-04-17 | Fusion-Io, Inc. | Apparatus, system, and method for validating that a correct data segment is read from a data storage device |
US8090901B2 (en) | 2009-05-14 | 2012-01-03 | Brocade Communications Systems, Inc. | TCAM management approach that minimize movements |
US9166818B2 (en) | 2009-09-21 | 2015-10-20 | Brocade Communications Systems, Inc. | Provisioning single or multistage networks using ethernet service instances (ESIs) |
US8599850B2 (en) | 2009-09-21 | 2013-12-03 | Brocade Communications Systems, Inc. | Provisioning single or multistage networks using ethernet service instances (ESIs) |
US20110216769A1 (en) * | 2010-03-08 | 2011-09-08 | Brocade Communications Systems, Inc. | Dynamic Path Selection |
WO2013040335A1 (en) * | 2011-09-14 | 2013-03-21 | Apple Inc. | Efficient non-volatile read cache for storage system |
US9569369B2 (en) * | 2011-10-27 | 2017-02-14 | Oracle International Corporation | Software translation lookaside buffer for persistent pointer management |
US20130111151A1 (en) * | 2011-10-27 | 2013-05-02 | Oracle International Corporation | Software translation lookaside buffer for persistent pointer management |
US9343177B2 (en) * | 2013-02-01 | 2016-05-17 | Apple Inc. | Accessing control registers over a data bus |
US20140223090A1 (en) * | 2013-02-01 | 2014-08-07 | Apple Inc | Accessing control registers over a data bus |
US9959044B2 (en) * | 2016-05-03 | 2018-05-01 | Macronix International Co., Ltd. | Memory device including risky mapping table and controlling method thereof |
US10545877B2 (en) * | 2018-04-05 | 2020-01-28 | Arm Limited | Apparatus and method for accessing an address translation cache |
US11055232B2 (en) | 2019-03-29 | 2021-07-06 | Intel Corporation | Valid bits of a translation lookaside buffer (TLB) for checking multiple page sizes in one probe cycle and reconfigurable sub-TLBS |
EP3719656A1 (en) * | 2019-03-29 | 2020-10-07 | Intel Corporation | Valid bits of a translation lookaside buffer (tlb) for checking multiple page sizes in one probe cycle and reconfigurable sub-tlbs |
Also Published As
Publication number | Publication date |
---|---|
KR101145557B1 (en) | 2012-05-16 |
US20120066475A1 (en) | 2012-03-15 |
CN1682200A (en) | 2005-10-12 |
AU2003260822A1 (en) | 2004-04-30 |
EP1552397A1 (en) | 2005-07-13 |
JP2005538465A (en) | 2005-12-15 |
CN100555248C (en) | 2009-10-28 |
KR20050043944A (en) | 2005-05-11 |
WO2004025480A1 (en) | 2004-03-25 |
US8607026B2 (en) | 2013-12-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8607026B2 (en) | Translation lookaside buffer | |
US5526504A (en) | Variable page size translation lookaside buffer | |
US6014732A (en) | Cache memory with reduced access time | |
US8984254B2 (en) | Techniques for utilizing translation lookaside buffer entry numbers to improve processor performance | |
US6732238B1 (en) | Set-associative cache memory having variable time decay rewriting algorithm | |
US7461208B1 (en) | Circuitry and method for accessing an associative cache with parallel determination of data and data availability | |
US4587610A (en) | Address translation systems for high speed computer memories | |
US7805588B2 (en) | Caching memory attribute indicators with cached memory data field | |
US5148538A (en) | Translation look ahead based cache access | |
EP0694845B1 (en) | Low-latency memory indexing method and structure | |
US5604879A (en) | Single array address translator with segment and page invalidate ability and method of operation | |
JP3666689B2 (en) | Virtual address translation method | |
US6874077B2 (en) | Parallel distributed function translation lookaside buffer | |
US6493812B1 (en) | Apparatus and method for virtual address aliasing and multiple page size support in a computer system having a prevalidated cache | |
US6138225A (en) | Address translation system having first and second translation look aside buffers | |
JPH07200399A (en) | Microprocessor and method for access to memory in microprocessor | |
US6581140B1 (en) | Method and apparatus for improving access time in set-associative cache systems | |
JPH04320553A (en) | Address converting mechanism | |
GB2293672A (en) | Virtual page memory buffer | |
EP0676698A1 (en) | Address translation circuit | |
US5535351A (en) | Address translator with by-pass circuit and method of operation | |
US20060047883A1 (en) | Serially indexing a cache memory | |
US5860097A (en) | Associative cache memory with improved hit time | |
US6385696B1 (en) | Embedded cache with way size bigger than page size | |
US6327646B1 (en) | Translation look-aside buffer utilizing high-order bits for fast access |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: KONINKLIJKE PHILIPS ELECTRONICS N.V., NETHERLANDS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:STRAVERS, PAUL;VAN DE WAERDT, JAN-WILLEM;REEL/FRAME:013344/0760;SIGNING DATES FROM 20020822 TO 20020917 |
|
AS | Assignment |
Owner name: NXP B.V., NETHERLANDS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:KONINKLIJKE PHILIPS ELECTRONICS N.V.;U.S. PHILIPS CORPORATION;REEL/FRAME:020548/0434 Effective date: 20070919 |
|
AS | Assignment |
Owner name: NXP B.V., NETHERLANDS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:KONINKLIJKE PHILIPS ELECTRONICS N.V.;REEL/FRAME:021085/0959 Effective date: 20080423 Owner name: NXP B.V.,NETHERLANDS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:KONINKLIJKE PHILIPS ELECTRONICS N.V.;REEL/FRAME:021085/0959 Effective date: 20080423 |
|
AS | Assignment |
Owner name: NXP B.V., NETHERLANDS Free format text: CHANGE OF NAME;ASSIGNOR:PHILIPS SEMICONDUCTORS INTERNATIONAL B.V.;REEL/FRAME:026233/0884 Effective date: 20060929 |
|
AS | Assignment |
Owner name: NXP B.V., NETHERLANDS Free format text: CONFIRMATORY ASSIGNMENT;ASSIGNORS:U.S. PHILIPS CORPORATION;KONINKLIJKE PHILIPS ELECTRONICS N.V.;SIGNING DATES FROM 20110531 TO 20110608;REEL/FRAME:026409/0628 |
|
AS | Assignment |
Owner name: KONINKLIJKE PHILIPS ELECTRONICS N.V., NETHERLANDS Free format text: CONFIRMATORY ASSIGNMENT;ASSIGNORS:STRAVERS, PAUL;VAN DE WAERDT, JAN-WILLEM;SIGNING DATES FROM 20110606 TO 20110608;REEL/FRAME:026425/0436 |
|
AS | Assignment |
Owner name: NYTELL SOFTWARE LLC, DELAWARE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:NXP B.V.;REEL/FRAME:026615/0576 Effective date: 20110628 |
|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- AFTER EXAMINER'S ANSWER OR BOARD OF APPEALS DECISION |