{"id":25033,"url":"https://patchwork.libcamera.org/api/1.1/patches/25033/?format=json","web_url":"https://patchwork.libcamera.org/patch/25033/","project":{"id":1,"url":"https://patchwork.libcamera.org/api/1.1/projects/1/?format=json","name":"libcamera","link_name":"libcamera","list_id":"libcamera_core","list_email":"libcamera-devel@lists.libcamera.org","web_url":"","scm_url":"","webscm_url":""},"msgid":"<20251114005428.90024-4-kieran.bingham@ideasonboard.com>","date":"2025-11-14T00:54:07","name":"[v4,03/21] ipa: libipa: Provide fixed point quantized traits","commit_ref":null,"pull_url":null,"state":"superseded","archived":false,"hash":"d9c5c7cd9da7bf3067684b257095b659d5b783cb","submitter":{"id":4,"url":"https://patchwork.libcamera.org/api/1.1/people/4/?format=json","name":"Kieran Bingham","email":"kieran.bingham@ideasonboard.com"},"delegate":null,"mbox":"https://patchwork.libcamera.org/patch/25033/mbox/","series":[{"id":5589,"url":"https://patchwork.libcamera.org/api/1.1/series/5589/?format=json","web_url":"https://patchwork.libcamera.org/project/libcamera/list/?series=5589","date":"2025-11-14T00:54:04","name":"libipa: Introduce a Quantized type","version":4,"mbox":"https://patchwork.libcamera.org/series/5589/mbox/"}],"comments":"https://patchwork.libcamera.org/api/patches/25033/comments/","check":"pending","checks":"https://patchwork.libcamera.org/api/patches/25033/checks/","tags":{},"headers":{"Return-Path":"","X-Original-To":"parsemail@patchwork.libcamera.org","Delivered-To":"parsemail@patchwork.libcamera.org","Received":["from lancelot.ideasonboard.com (lancelot.ideasonboard.com\n\t[92.243.16.209])\n\tby patchwork.libcamera.org (Postfix) with ESMTPS id 81673C3263\n\tfor ;\n\tFri, 14 Nov 2025 00:54:58 +0000 (UTC)","from lancelot.ideasonboard.com (localhost [IPv6:::1])\n\tby lancelot.ideasonboard.com (Postfix) with ESMTP id EFEA960A8A;\n\tFri, 14 Nov 2025 01:54:52 +0100 (CET)","from perceval.ideasonboard.com (perceval.ideasonboard.com\n\t[IPv6:2001:4b98:dc2:55:216:3eff:fef7:d647])\n\tby lancelot.ideasonboard.com (Postfix) with ESMTPS id 2B62B6069A\n\tfor ;\n\tFri, 14 Nov 2025 01:54:49 +0100 (CET)","from charm.hippo-penny.ts.net\n\t(cpc89244-aztw30-2-0-cust6594.18-1.cable.virginm.net [86.31.185.195])\n\tby perceval.ideasonboard.com (Postfix) with ESMTPSA id 9774819AD;\n\tFri, 14 Nov 2025 01:52:48 +0100 (CET)"],"Authentication-Results":"lancelot.ideasonboard.com; dkim=pass (1024-bit key;\n\tunprotected) header.d=ideasonboard.com header.i=@ideasonboard.com\n\theader.b=\"sQKqjYXF\"; dkim-atps=neutral","DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/simple; d=ideasonboard.com;\n\ts=mail; t=1763081568;\n\tbh=iI4pJjXoTff8rWtLHaq/vvewmwwoZYk8584XcCi+8GU=;\n\th=From:To:Cc:Subject:Date:In-Reply-To:References:From;\n\tb=sQKqjYXFCjV/Nn+rvGyGy0gW6H1CErkNKkaYbmINaTKrhtJaCqBq+4Ki/wDfy6Irl\n\tC40K1hM6G1sV4clMwzaBm5PZJcI45KXwXEmWhpWH2BfsDFwXtJ8KyYOn533Kvcd5VX\n\tJKL7GBGjP90F9/qmcX8KaG7+Bd1rJplun0r5fIek=","From":"Kieran Bingham ","To":"libcamera devel ","Cc":"Kieran Bingham ","Subject":"[PATCH v4 03/21] ipa: libipa: Provide fixed point quantized traits","Date":"Fri, 14 Nov 2025 00:54:07 +0000","Message-ID":"<20251114005428.90024-4-kieran.bingham@ideasonboard.com>","X-Mailer":"git-send-email 2.51.1","In-Reply-To":"<20251114005428.90024-1-kieran.bingham@ideasonboard.com>","References":"<20251114005428.90024-1-kieran.bingham@ideasonboard.com>","MIME-Version":"1.0","Content-Transfer-Encoding":"8bit","X-BeenThere":"libcamera-devel@lists.libcamera.org","X-Mailman-Version":"2.1.29","Precedence":"list","List-Id":"","List-Unsubscribe":",\n\t","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n\t","Errors-To":"libcamera-devel-bounces@lists.libcamera.org","Sender":"\"libcamera-devel\" "},"content":"Extend the new Quantized type infrastructure by providing a\nFixedPointQTraits template.\n\nThis allows construction of fixed point types with a Quantized storage\nthat allows easy reading of both the underlying quantized type value and\na floating point representation of that same value.\n\nSigned-off-by: Kieran Bingham \n\n---\nv4:\n- Assert that the given type has enough bits for the usage\n- Use unsigned types for calculating qmin/qmax\n- Reorder toFloat/fromFloat and min/max for future inlining\n- Make toFloat and fromFloat constexpr\n\n src/ipa/libipa/fixedpoint.cpp | 121 ++++++++++++++++++++++++++++++++++\n src/ipa/libipa/fixedpoint.h | 45 +++++++++++++\n 2 files changed, 166 insertions(+)","diff":"diff --git a/src/ipa/libipa/fixedpoint.cpp b/src/ipa/libipa/fixedpoint.cpp\nindex 6b698fc5d680..4c3ea33497bf 100644\n--- a/src/ipa/libipa/fixedpoint.cpp\n+++ b/src/ipa/libipa/fixedpoint.cpp\n@@ -37,6 +37,127 @@ namespace ipa {\n * \\return The converted value\n */\n \n+/**\n+ * \\struct libcamera::ipa::FixedPointQTraits\n+ * \\brief Traits type implementing fixed-point quantisation conversions\n+ *\n+ * The FixedPointQTraits structure defines a policy for mapping floating-point\n+ * values to and from fixed-point integer representations. It is parameterised\n+ * by the number of integer bits \\a I, fractional bits \\a F, and the integral\n+ * storage type \\a T. The traits are used with Quantized to create a\n+ * quantised type that stores both the fixed-point representation and the\n+ * corresponding floating-point value.\n+ *\n+ * The trait exposes compile-time constants describing the bit layout, limits,\n+ * and scaling factors used in the fixed-point representation.\n+ *\n+ * \\tparam I Number of integer bits\n+ * \\tparam F Number of fractional bits\n+ * \\tparam T Integral type used to store the quantised value\n+ */\n+\n+/**\n+ * \\typedef FixedPointQTraits::QuantizedType\n+ * \\brief The integral storage type used for the fixed-point representation\n+ */\n+\n+/**\n+ * \\typedef FixedPointQTraits::UT\n+ * \\brief The unsigned representation of the integral storage\n+ *\n+ * This type definition is to support internal usage only.\n+ */\n+\n+/**\n+ * \\var FixedPointQTraits::Bits\n+ * \\brief Total number of bits used in the fixed-point format (I + F)\n+ */\n+\n+/**\n+ * \\var FixedPointQTraits::BitMask\n+ * \\brief Bit mask selecting all valid bits in the fixed-point representation\n+ */\n+\n+/**\n+ * \\var FixedPointQTraits::qmin\n+ * \\brief Minimum representable quantised integer value\n+ *\n+ * This corresponds to the most negative value for signed formats or zero for\n+ * unsigned formats.\n+ */\n+\n+/**\n+ * \\var FixedPointQTraits::qmax\n+ * \\brief Maximum representable quantised integer value\n+ */\n+\n+/**\n+ * \\var FixedPointQTraits::min\n+ * \\brief Minimum representable floating-point value corresponding to qmin\n+ */\n+\n+/**\n+ * \\var FixedPointQTraits::max\n+ * \\brief Maximum representable floating-point value corresponding to qmax\n+ */\n+\n+/**\n+ * \\fn FixedPointQTraits::fromFloat(float v)\n+ * \\brief Convert a floating-point value to a fixed-point integer\n+ * \\param[in] v The floating-point value to be converted\n+ * \\return The quantised fixed-point integer representation\n+ *\n+ * The conversion rounds the floating-point input \\a v to the nearest integer\n+ * according to the scaling factor defined by the number of fractional bits F.\n+ */\n+\n+/**\n+ * \\fn FixedPointQTraits::toFloat(QuantizedType q)\n+ * \\brief Convert a fixed-point integer to a floating-point value\n+ * \\param[in] q The fixed-point integer value to be converted\n+ * \\return The corresponding floating-point value\n+ *\n+ * The conversion sign-extends the integer value if required and divides by the\n+ * scaling factor defined by the number of fractional bits F.\n+ */\n+\n+/**\n+ * \\typedef Q1_7\n+ * \\brief 1.7 signed fixed-point quantizer\n+ *\n+ * A Quantized type using 1 bit for the integer part and 7 bits for the\n+ * fractional part, stored in a signed 8-bit integer (\\c int8_t). Represents\n+ * values approximately in the range [-1.0, 0.992] with a resolution of 1/128.\n+ */\n+\n+/**\n+ * \\typedef UQ1_7\n+ * \\brief 1.7 unsigned fixed-point quantizer\n+ *\n+ * A Quantized type using 1 bit for the integer part and 7 bits for the\n+ * fractional part, stored in an unsigned 8-bit integer (\\c uint8_t). Represents\n+ * values in the range [0.0, 1.992] with a resolution of 1/128.\n+ */\n+\n+/**\n+ * \\typedef Q12_4\n+ * \\brief 12.4 signed fixed-point quantizer\n+ *\n+ * A Quantized type using 12 bits for the integer part and 4 bits for the\n+ * fractional part, stored in a signed 16-bit integer (\\c int16_t). Represents\n+ * values in the range approximately [-2048.0, 2047.9375] with a resolution of\n+ * 1/16.\n+ */\n+\n+/**\n+ * \\typedef UQ12_4\n+ * \\brief 12.4 unsigned fixed-point quantizer\n+ *\n+ * A Quantized type using 12 bits for the integer part and 4 bits for the\n+ * fractional part, stored in an unsigned 16-bit integer (\\c uint16_t).\n+ * Represents values in the range [0.0, 4095.9375] with a resolution of 1/16.\n+ */\n+\n } /* namespace ipa */\n \n } /* namespace libcamera */\ndiff --git a/src/ipa/libipa/fixedpoint.h b/src/ipa/libipa/fixedpoint.h\nindex 709cf50f0fcd..17db8a026502 100644\n--- a/src/ipa/libipa/fixedpoint.h\n+++ b/src/ipa/libipa/fixedpoint.h\n@@ -10,6 +10,8 @@\n #include \n #include \n \n+#include \"quantized.h\"\n+\n namespace libcamera {\n \n namespace ipa {\n@@ -60,6 +62,49 @@ constexpr R fixedToFloatingPoint(T number)\n \treturn static_cast(t) / static_cast(1 << F);\n }\n \n+template\n+struct FixedPointQTraits {\n+\tstatic_assert(std::is_integral_v, \"FixedPointQTraits: T must be integral\");\n+\tusing QuantizedType = T;\n+\tusing UT = std::make_unsigned_t;\n+\n+\tstatic constexpr unsigned int Bits = I + F;\n+\tstatic_assert(Bits <= sizeof(T) * 8, \"FixedPointQTraits: too many bits for type T\");\n+\n+\tstatic constexpr T BitMask = (Bits < sizeof(T) * 8)\n+\t\t\t\t ? static_cast((UT{1} << Bits) - 1)\n+\t\t\t\t : static_cast(~UT{0});\n+\n+\tstatic constexpr T qmin = std::is_signed_v\n+\t\t\t\t? static_cast(-(UT{1} << (Bits - 1)))\n+\t\t\t\t: static_cast(0);\n+\n+\tstatic constexpr T qmax = std::is_signed_v\n+\t\t\t\t? static_cast((UT{1} << (Bits - 1)) - 1)\n+\t\t\t\t: static_cast((UT{1} << Bits) - 1);\n+\n+\tstatic constexpr float toFloat(QuantizedType q)\n+\t{\n+\t\treturn fixedToFloatingPoint(q);\n+\t}\n+\n+\tstatic constexpr float min = fixedToFloatingPoint(qmin);\n+\tstatic constexpr float max = fixedToFloatingPoint(qmax);\n+\n+\t/* Conversion functions required by Quantized */\n+\tstatic constexpr QuantizedType fromFloat(float v)\n+\t{\n+\t\tv = std::clamp(v, min, max);\n+\t\treturn floatingToFixedPoint(v);\n+\t}\n+};\n+\n+using Q1_7 = Quantized>;\n+using UQ1_7 = Quantized>;\n+\n+using Q12_4 = Quantized>;\n+using UQ12_4 = Quantized>;\n+\n } /* namespace ipa */\n \n } /* namespace libcamera */\n","prefixes":["v4","03/21"]}