[libcamera-devel] [PATCH 1/6] ipa: raspberrypi: Non-functional formatting fixes to md_parser.hpp
Laurent Pinchart
laurent.pinchart at ideasonboard.com
Tue Jun 15 04:36:50 CEST 2021
On Tue, Jun 15, 2021 at 05:35:42AM +0300, Laurent Pinchart wrote:
> On Mon, Jun 14, 2021 at 10:53:35AM +0100, Naushir Patuck wrote:
> > Adjust source formatting to closer match libcamera guidelines:
> >
> > - Remove unused header files.
> > - Switch to C style comments.
> > - Add whitespace for readability.
> >
> > There are no functional changes in this commit.
> >
> > Signed-off-by: Naushir Patuck <naush at raspberrypi.com>
> > ---
> > src/ipa/raspberrypi/md_parser.hpp | 144 ++++++++++++++++++------------
> > 1 file changed, 85 insertions(+), 59 deletions(-)
> >
> > diff --git a/src/ipa/raspberrypi/md_parser.hpp b/src/ipa/raspberrypi/md_parser.hpp
> > index 8e22b1d7ca43..86e0577614e0 100644
> > --- a/src/ipa/raspberrypi/md_parser.hpp
> > +++ b/src/ipa/raspberrypi/md_parser.hpp
> > @@ -6,75 +6,94 @@
> > */
> > #pragma once
> >
> > -#include <stdint.h>
> > -
Oh, and this should be kept, as the header uses uint*_t types.
> > #include <libcamera/span.h>
> >
> > /* Camera metadata parser class. Usage as shown below.
>
> This should be
>
> /*
> * Camera metadata parser class. Usage as shown below.
>
> I'll fix when applying.
>
> Reviewed-by: Laurent Pinchart <laurent.pinchart at ideasonboard.com>
>
> > -
> > -Setup:
> > -
> > -Usually the metadata parser will be made as part of the CamHelper class so
> > -application code doesn't have to worry which to kind to instantiate. But for
> > -the sake of example let's suppose we're parsing imx219 metadata.
> > -
> > -MdParser *parser = new MdParserImx219(); // for example
> > -parser->SetBitsPerPixel(bpp);
> > -parser->SetLineLengthBytes(pitch);
> > -parser->SetNumLines(2);
> > -
> > -Note 1: if you don't know how many lines there are, the size of the input
> > -buffer is used as a limit instead.
> > -
> > -Note 2: if you don't know the line length, you can leave the line length unset
> > -(or set to zero) and the parser will hunt for the line start instead.
> > -
> > -Then on every frame:
> > -
> > -if (parser->Parse(buffer) != MdParser::OK)
> > - much badness;
> > -unsigned int exposure_lines, gain_code
> > -if (parser->GetExposureLines(exposure_lines) != MdParser::OK)
> > - exposure was not found;
> > -if (parser->GetGainCode(parser, gain_code) != MdParser::OK)
> > - gain code was not found;
> > -
> > -(Note that the CamHelper class converts to/from exposure lines and time,
> > -and gain_code / actual gain.)
> > -
> > -If you suspect your embedded data may have changed its layout, change any line
> > -lengths, number of lines, bits per pixel etc. that are different, and
> > -then:
> > -
> > -parser->Reset();
> > -
> > -before calling Parse again. */
> > + *
> > + * Setup:
> > + *
> > + * Usually the metadata parser will be made as part of the CamHelper class so
> > + * application code doesn't have to worry which to kind to instantiate. But for
> > + * the sake of example let's suppose we're parsing imx219 metadata.
> > + *
> > + * MdParser *parser = new MdParserImx219(); // for example
> > + * parser->SetBitsPerPixel(bpp);
> > + * parser->SetLineLengthBytes(pitch);
> > + * parser->SetNumLines(2);
> > + *
> > + * Note 1: if you don't know how many lines there are, the size of the input
> > + * buffer is used as a limit instead.
> > + *
> > + * Note 2: if you don't know the line length, you can leave the line length unset
> > + * (or set to zero) and the parser will hunt for the line start instead.
> > + *
> > + * Then on every frame:
> > + *
> > + * if (parser->Parse(buffer) != MdParser::OK)
> > + * much badness;
> > + * unsigned int exposure_lines, gain_code
> > + * if (parser->GetExposureLines(exposure_lines) != MdParser::OK)
> > + * exposure was not found;
> > + * if (parser->GetGainCode(parser, gain_code) != MdParser::OK)
> > + * gain code was not found;
> > + *
> > + * (Note that the CamHelper class converts to/from exposure lines and time,
> > + * and gain_code / actual gain.)
> > + *
> > + * If you suspect your embedded data may have changed its layout, change any line
> > + * lengths, number of lines, bits per pixel etc. that are different, and
> > + * then:
> > + *
> > + * parser->Reset();
> > + *
> > + * before calling Parse again.
> > + */
> >
> > namespace RPiController {
> >
> > -// Abstract base class from which other metadata parsers are derived.
> > +/* Abstract base class from which other metadata parsers are derived. */
> >
> > class MdParser
> > {
> > public:
> > - // Parser status codes:
> > - // OK - success
> > - // NOTFOUND - value such as exposure or gain was not found
> > - // ERROR - all other errors
> > + /*
> > + * Parser status codes:
> > + * OK - success
> > + * NOTFOUND - value such as exposure or gain was not found
> > + * ERROR - all other errors
> > + */
> > enum Status {
> > OK = 0,
> > NOTFOUND = 1,
> > ERROR = 2
> > };
> > - MdParser() : reset_(true) {}
> > +
> > + MdParser() : reset_(true)
> > + {
> > + }
> > +
> > virtual ~MdParser() = default;
> > - void Reset() { reset_ = true; }
> > - void SetBitsPerPixel(int bpp) { bits_per_pixel_ = bpp; }
> > - void SetNumLines(unsigned int num_lines) { num_lines_ = num_lines; }
> > +
> > + void Reset()
> > + {
> > + reset_ = true;
> > + }
> > +
> > + void SetBitsPerPixel(int bpp)
> > + {
> > + bits_per_pixel_ = bpp;
> > + }
> > +
> > + void SetNumLines(unsigned int num_lines)
> > + {
> > + num_lines_ = num_lines;
> > + }
> > +
> > void SetLineLengthBytes(unsigned int num_bytes)
> > {
> > line_length_bytes_ = num_bytes;
> > }
> > +
> > virtual Status Parse(libcamera::Span<const uint8_t> buffer) = 0;
> > virtual Status GetExposureLines(unsigned int &lines) = 0;
> > virtual Status GetGainCode(unsigned int &gain_code) = 0;
> > @@ -87,22 +106,28 @@ protected:
> > unsigned int buffer_size_bytes_;
> > };
> >
> > -// This isn't a full implementation of a metadata parser for SMIA sensors,
> > -// however, it does provide the findRegs method which will prove useful and make
> > -// it easier to implement parsers for other SMIA-like sensors (see
> > -// md_parser_imx219.cpp for an example).
> > +/*
> > + * This isn't a full implementation of a metadata parser for SMIA sensors,
> > + * however, it does provide the findRegs method which will prove useful and make
> > + * it easier to implement parsers for other SMIA-like sensors (see
> > + * md_parser_imx219.cpp for an example).
> > + */
> >
> > class MdParserSmia : public MdParser
> > {
> > public:
> > - MdParserSmia() : MdParser() {}
> > + MdParserSmia() : MdParser()
> > + {
> > + }
> >
> > protected:
> > - // Note that error codes > 0 are regarded as non-fatal; codes < 0
> > - // indicate a bad data buffer. Status codes are:
> > - // PARSE_OK - found all registers, much happiness
> > - // MISSING_REGS - some registers found; should this be a hard error?
> > - // The remaining codes are all hard errors.
> > + /*
> > + * Note that error codes > 0 are regarded as non-fatal; codes < 0
> > + * indicate a bad data buffer. Status codes are:
> > + * PARSE_OK - found all registers, much happiness
> > + * MISSING_REGS - some registers found; should this be a hard error?
> > + * The remaining codes are all hard errors.
> > + */
> > enum ParseStatus {
> > PARSE_OK = 0,
> > MISSING_REGS = 1,
> > @@ -112,6 +137,7 @@ protected:
> > BAD_LINE_END = -4,
> > BAD_PADDING = -5
> > };
> > +
> > ParseStatus findRegs(libcamera::Span<const uint8_t> buffer, uint32_t regs[],
> > int offsets[], unsigned int num_regs);
> > };
--
Regards,
Laurent Pinchart
More information about the libcamera-devel
mailing list