This patent document contains information subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent, as it appears in the US Patent and Trademark Office files or records, but otherwise reserves all copyright rights whatsoever.
The present disclosure relates to memory structures for transferring image frames to a display of an embedded device. Aspects of this disclosure relate to providing, on an image display processor chip of an embedded device, an interim buffer from which pixels are rendered to a display device.
Various types of embedded devices include a main memory portion, which holds image frames. The image frames held by the frame buffer in the main memory are sent to a display via an image display processor chip. The image display processor chip may include a mechanism to manipulate the positions of pixels within the image before the image is sent to the display. For example, the image may be rotated. This may be done by manipulating the pixels, i.e., moving the pixels, and storing the resulting manipulated image in a buffer within the main memory before sending the manipulated pixels to the display. Alternatively, or in addition, the resulting manipulated image may be stored within a buffer that is part of the display processor chip.
Embedded systems, for example, mobile phones, have limited memory resources. A given embedded system may have a main memory and a system bus, both of which are shared by different hardware entities, including an image processing chip. Meanwhile, the embedded system image processing chip may require large amounts of bandwidth of the main memory via the system bus.
Memory bandwidth demands like this can result in a memory access bottleneck, which could delay the operation of the video processing chip as well as other hardware entities that use the same main memory and system bus.
To reduce this bottleneck, on-chip memory may be provided within the image processing chip, and the image processing chip may place the manipulated image in this on-chip memory. However, increasing the amount of memory within the image processing chip consumes the resources of the chip, possibly also increasing the cost of the chip.
Apparatus are provided, including an embedded display processor on a given chip. The apparatus may be an embedded device, for example, a mobile wireless communications device. More specifically, the apparatus may be a mobile phone, a portable gaming device, a video streaming device, or a GPS map drawing device.
The display processor includes, on the same given chip, a rendering memory, from which pixels are rendered to a display device.
The display processor further includes an image manipulation mechanism to manipulate pixels of a given image frame from source positions in a pre-manipulation buffer, to target positions in the rendering memory, the target positions corresponding to rendered positions in the given image frame.
The display processor further includes a fetch mechanism to fetch, from the pre-manipulation buffer, a predetermined number of neighboring pixels including adjacent cross-row pixels traversing a plurality of rows while in their source positions, the adjacent cross-row pixels being intended for target positions in a common row in accordance with the manipulation to be performed by the manipulation mechanism.
The display processor further includes a send mechanism to send, from the rendering memory, a set of the neighboring pixels to the display device in accordance with a given dynamic refresh rate and scheme. The set of the neighboring pixels include adjacent common row pixels on a common row after having been manipulated.
The display processor further includes a reconfigure mechanism to periodically reconfigure a manner of assignment of addresses and physical locations for data stored in the rendering memory.
Embodiments of the disclosure are further described in the detailed description, which follows, by reference to the noted drawings, in which like reference numerals represents similar parts throughout the several views of the drawings, and wherein:
Referring now to the drawings in greater detail,
Mobile station modem chip 16 further includes, in addition to video codec 18, a memory controller 20 and a mobile display processor 21. The illustrated mobile display processor 21 includes a buffer 22, an image manipulation mechanism 24, an upscale and color convert mechanism 26, a virtual device buffer 28, a send mechanism 29, and other display processing elements 30 (for example, a 3-D graphics pipeline).
The illustrated mobile display processor 21 may be part of a larger integrated circuit or chip, it may be a separate processor chip, or it may be part of an overall system on a chip (SoC). The display processor 21 includes, on the same chip, a rendering memory, which is a virtual device buffer (VDB) 28 in the embodiment shown in
The image manipulation mechanism 24 may include a fetch, flip and rotate (FFR) mechanism to reposition pixels from source positions in a pre-manipulation buffer (application buffer(s) 14 in the illustrated embodiment). The source positions correspond to positions in a given image. The pixels are repositioned as they are moved to target positions in the rendering memory (virtual device buffer 28 in the illustrated embodiment). That manipulation may include, for example, rotation of the image.
The illustrated display processor 21 further includes a fetch mechanism to fetch, from the pre-manipulation buffer (application buffer(s) 14 as illustrated) a predetermined number of neighboring pixels. In the illustrated embodiment, the fetch mechanism includes memory controller 20.
The neighboring pixels that are fetched include adjacent cross-row pixels traversing a plurality of rows while in their source positions. The adjacent cross-row pixels may be intended for target positions in a common row in accordance with the manipulation to be performed by the manipulation mechanism. For example, if the manipulation mechanism is performing a rotation on the image, the adjacent cross-row pixels being fetched may end up, for example, as common-row pixels when the rotation is a clockwise rotation of 90 degrees.
A send mechanism 29 may be provided, to send, from the rendering memory, a predetermined number of neighboring pixels, including adjacent common-row pixels on a common row after having been manipulated, in the case where, for example, the image was rotated in one direction by 90 degrees.
The neighboring pixels are sent to the display device in accordance with a given dynamic refresh rate and scheme. In the illustrated embodiment, send mechanism 29 includes features to send data using a follow-the-beam approach, whereby the contents of the frame buffer (in this case the rendering memory) are updated as the content for the same location within the frame has been sent to the display. The send mechanism may be part of the mobile display processor chip, which is also part of the same mobile station modem chip 16, in the illustrated embodiment.
The illustrated mobile display processor 21 may further include a reconfigure mechanism 31 to periodically reconfigure the manner in which addresses and physical locations are allocated to the data stored in the rendering memory (virtual device buffer 28) as room within the rendering memory becomes available for additional data.
In act 44, the pixels from the given plural pixel region are written into the target pixel locations. In the illustrated embodiment, the target pixel locations are locations within a rendering memory, which specifically is virtual device buffer 28.
As the source pixels are processed, before they are stored in the rendering memory, the pixel data is translated to an appropriate device space corresponding to the ultimate display device to be used to display the image.
In act 44, the process of writing pixels into the target locations includes manipulating the source pixels of the given region, which may include, for example, rotating or flipping of those pixels before placing them into the target locations.
As shown in act 40 in
In act 50, the source pixels of a given region within the source image are manipulated. For example, they are rotated and/or flipped.
In the example shown in
In act 52, in
In act 54, space is reserved for a given line in the interim target memory. The reserved lines span the width of the display screen (in pixels). In act 56, space is reserved for consecutive adjacent lines adjacent to the given line, resulting in a total reserved space for lines 1-N. N is equal to the height of the blocks. In the illustrated embodiment, a block is four pixels high and four pixels wide, in terms of the pixel positions as manipulated and as stored in the interim target memory. Addresses are logically assigned to the reserved space, so the reserved space can be filled in the correct order, and so that the data can be sent in consecutive lines in a manner that is consistent with a follow-the-beam process.
In act 58, for the given region (a four by four pixel block in the illustrated embodiment) the top slice of that region (slice 0) is written to the reserved given line, and subsequent slices (1, 2, 3) are written to the respective subsequent reserved adjacent lines.
In act 60, which is shown at the top of
In a next act 64, as the physical space in the interim target for the given line becomes available (i.e., because its data has been sent to the display), it is reassociated with new addresses corresponding to the next region to be transferred from the source memory to the interim target. This is explained in an example below with reference to
In act 64, as more reserved space for the given line becomes available, it is moved logically and associated with a next region to be transferred from the source memory.
At act 66, which is shown at the top of
In the example illustrated in
As noted above, as individual lines becomes free, i.e., the data for those lines have already been sent to the display, within the virtual device buffer, the slice locations for those lines can be reassociated with unique addresses so that they can accept data for subsequent source pixels.
As shown in
After five blocks are written into the example simplified interim target memory, it is now full, and the first line is available to be sent to the display. In other words, a given line in the interim target memory is not sent to the interim target, in the illustrated embodiment, until it is completely filled with data from the source. Since the data is written one block at a time, in the illustrated example, this requires that the first five blocks will have been written into the interim target memory. Once the read pointer (not shown) has passed line 1, line 1 is sent to the display.
A block, in this example, requires 4 slices, and now (after line 1 has been sent to the display) 5 empty slices are available, specifically, slices 0, 4, 8, 12, and 16 from which data has been sent to the display. Accordingly, the virtual address assignment for the physical spaces associated with these slices will remain the same, but those same addresses will be reserved to accept incoming data read from the source memory in a different order. The order in which the data is now accepted for a next block from the source memory is dictated by the positions at which these addresses are now located in
In the illustrated embodiment, slice addresses 0, 4, 8, and 12 still correspond to the same physical locations within the memory, but they now correspond to the block data in a different way. Whereas, previously, slice locations 0, 1, 2, and 3 corresponded to a common block, now slice locations 0, 4, 8, and 12, incremented by 4, correspond to a next common block.
As shown in
As shown in
As shown in
The reconfiguration scheme illustrated in
The increment value is the number of slices per block raised to the nth (where n is the stripe number) modulo 19. These numbers can be computed by taking a previous increment value and multiplying that by the slices per block, and applying that result to the modulo function.
The block writing address order in this scheme shown in
Note that the bottom stripe in the reconfiguration scheme as shown in
The claims as originally presented and as they may be amended, encompass variations, alternatives, modifications, improvements, equivalents and substantial equivalents of the embodiments and teachings disclosed herein, including those that are presently unforeseen or unappreciated, and that, for example, may arise from applicants/patentees and others.
This application claims the benefit of provisional U.S. Application Ser. No. 60/696,187, entitled “VDB THEORY AND OPERATION” filed on Jun. 29, 2005, assigned to the assignee of the present application, and incorporated herein by reference in its entirety for all purposes.
Number | Name | Date | Kind |
---|---|---|---|
5111192 | Kadakia | May 1992 | A |
5412766 | Pietras et al. | May 1995 | A |
5598181 | Kermisch | Jan 1997 | A |
5801717 | Engstrom et al. | Sep 1998 | A |
5844560 | Crutcher et al. | Dec 1998 | A |
5936616 | Torborg et al. | Aug 1999 | A |
5966116 | Wakeland | Oct 1999 | A |
5986672 | Groezinger et al. | Nov 1999 | A |
6366289 | Johns | Apr 2002 | B1 |
6757447 | Yamaguchi et al. | Jun 2004 | B2 |
6825845 | Nally | Nov 2004 | B2 |
7050071 | Wyatt et al. | May 2006 | B2 |
7376286 | Locker et al. | May 2008 | B2 |
20040052431 | Locker et al. | Mar 2004 | A1 |
20040239690 | Wyatt et al. | Dec 2004 | A1 |
20050184993 | Ludwin et al. | Aug 2005 | A1 |
20050275665 | Kejser | Dec 2005 | A1 |
Number | Date | Country | |
---|---|---|---|
20070002060 A1 | Jan 2007 | US |
Number | Date | Country | |
---|---|---|---|
60696187 | Jun 2005 | US |