Utilisation de D3DXMath

D3DXMath est une bibliothèque d’assistance mathématique pour les applications Direct3D. D3DXMath est de longue date, est inclus dans D3DX 9 et D3DX 10, et remonte également aux versions antérieures de DirectX.

Notes

La bibliothèque d’utilitaireS D3DX (D3DX 9, D3DX 10 et D3DX 11) étant déconseillée pour Windows 8, nous vous recommandons vivement de migrer vers DirectXMath plutôt que d’utiliser D3DXMath.

 

DirectXMath partage une grande partie des mêmes fonctionnalités dans D3DXMath, et D3DXMath inclut en interne un certain nombre d’optimisations spécifiques au processeur. La principale différence est que D3DXMath est hébergé dans le D3DX9*. DLL et D3DX10*. Les DLL, et très peu de fonctions sont incluses. La convention d’appel de la bibliothèque DirectXMath est explicitement compatible AVEC SIMD, tandis que D3DXMath doit effectuer des conversions de charge et de stockage pour implémenter l’optimisation SIMD.

Mélange de DirectXMath et D3DXMath

D3DX11 ne contient pas D3DXMath et, en général, nous vous recommandons d’utiliser DirectXMath à la place. Toutefois, vous êtes libre de continuer à lier D3DX9 et/ou D3DX10 dans votre application. Par conséquent, vous pouvez continuer à utiliser D3DXMath ou utiliser D3DXMath et DirectXMath dans votre application en même temps.

Il est généralement sûr de caster un XMVECTOR* en fonction qui prend D3DXVECTOR4* ou de caster un XMMATRIX* en fonction qui prend D3DXMATRIX*. L’inverse n’est toutefois pas généralement sûr, car XMVECTOR et XMMATRIX doivent être alignés sur 16 octets, tandis que D3DXVECTOR4 et D3DXMATRIX n’ont aucune exigence de ce type. Le non-respect de cette exigence peut entraîner des exceptions d’alignement non valides au moment de l’exécution.

Il est sûr de caster un XMVECTOR* vers une fonction qui prend D3DXVECTOR2* ou D3DXVECTOR3*, mais pas inversement. Les deux problèmes d’alignement et le fait que D3DXVECTOR2 et D3DXVECTOR3 sont des structures plus petites font de cette opération une opération non sécurisée.

Notes

D3DX (et donc D3DXMath) est considéré comme hérité et n’est pas disponible pour les applications du Windows Store qui s’exécutent sur Windows 8 et n’est pas inclus dans le Kit de développement logiciel (SDK) Windows 8 pour les applications de bureau.

 

Utilisation de DirectXMath avec Direct3D

DirectXMath et D3DXMath sont facultatifs lors de l’utilisation de Direct3D. Direct3D 9 a défini D3DMATRIX et D3DCOLOR dans le cadre de l’API Direct3D pour prendre en charge le pipeline à fonction fixe (maintenant hérité). D3DXMath dans D3DX9 étend ces types Direct3D 9 avec des opérations mathématiques graphiques courantes. Pour Direct3D 10.x et Direct3D 11, l’API utilise uniquement le pipeline programmable, de sorte qu’il n’existe aucune structure spécifique à l’API pour les matrices ou les valeurs de couleur. Lorsque les API les plus récentes nécessitent une valeur de couleur, elles prennent un tableau explicite de valeurs float ou une mémoire tampon générique de données constantes interprétées par le nuanceur HLSL. HLSL lui-même peut prendre en charge les formats de matrice principale de ligne ou de colonne principale, de sorte que la disposition vous appartient entièrement (pour plus d’informations, consultez HLSL, Classement des matrices ; si vous utilisez des formats de matrice de colonne principale dans vos nuanceurs, vous devez transposer les données de matrice DirectXMath lorsque vous les placez dans vos structures de mémoire tampon constantes). Bien qu’elles soient facultatives, les bibliothèques DirectXMath et D3DXMath fournissent toutes deux des fonctionnalités graphiques courantes et sont donc extrêmement pratiques lors de la programmation Direct3D.

Il est sûr de convertir un XMVECTOR* en D3DVECTOR* ou XMMATRIX* en D3DMATRIX* car Direct3D 9 n’émet aucune hypothèse d’alignement sur la structure de données entrante. Il est également sûr de caster XMCOLOR en D3DCOLOR. Vous pouvez convertir une représentation 4-float de couleur en XMCOLOR via XMStoreColor() pour obtenir le DWORD 8:8:8:8:8 32 bits équivalent à D3DCOLOR.

Lorsque vous utilisez Direct3D 10.x ou Direct3D 11, vous utilisez généralement les types DirectXMath pour créer une structure pour chacune de vos mémoires tampons constantes, et dans ce cas, cela dépend en grande partie de votre capacité à contrôler l’alignement pour rendre ces opérations efficaces, ou à utiliser les opérations XMStore*() pour convertir les données XMVECTOR et XMMATRIX vers les types de données appropriés. Lorsque vous appelez des API Direct3D 10.x ou Direct3D 11 qui nécessitent un tableau float[4] de valeurs de couleur, vous pouvez caster un XMVECTOR* ou XMFLOAT4* contenant les données de couleur.

Portage à partir de D3DXMath

D3DXMath Type Équivalent DirectXMath
D3DXFLOAT16 MOITIÉ
D3DXMATRIX XMFLOAT4X4
D3DXMATRIXA16 XMMATRIX ou XMFLOAT4X4A
D3DXQUATERNION
D3DXPLANE
D3DXCOLOR
XMVECTOR est utilisé plutôt que d’avoir des types uniques. Vous devrez donc probablement utiliser un XMFLOAT4Remarque :**D3DXQUATERNION::operator *** appelle la fonction D3DXQuaternionMultiply , qui multiplie deux quaternions. Toutefois, à moins d’utiliser explicitement la fonction XMQuaternionMultiply , vous obtenez une réponse incorrecte lorsque vous utilisez **XMVECTOR::operator *** sur un quaternion.
D3DXVECTOR2 XMFLOAT2
D3DXVECTOR2_16F XMHALF2
D3DXVECTOR3 XMFLOAT3
D3DXVECTOR4 XMFLOAT4(ou si vous pouvez garantir que les données sont alignées sur 16 octets, XMVECTOR ou XMFLOAT4A )
D3DXVECTOR4_16F XMHALF4

 

Notes

Il n’existe aucun équivalent direct à D3DXVECTOR3_16F dans XNAMath.

 

D3DXMath Macro Équivalent DirectXMath
D3DX_PI XM_PI
D3DX_1BYPI XM_1DIVPI
D3DXToRadian XMConvertToRadians
D3DXToDegree XMConvertToDegrees

 

D3DXMath, fonction Équivalent DirectXMath
D3DXBoxBoundProbe BoundingBox::Intersects(XMVECTOR, XMVECTOR, float&)
D3DXComputeBoundingBox BoundingBox::CreateFromPoints
D3DXComputeBoundingSphere BoundingSphere::CreateFromPoints
D3DXSphereBoundProbe BoundingSphere::Intersects(XMVECTOR, XMVECTOR, float&)
D3DXIntersectTriFunction TriangleTests::Intersects
D3DXFloat32To16Array XMConvertFloatToHalfStream
D3DXFloat16To32Array XMConvertHalfToFloatStream
D3DXVec2Length XMVector2Length ou XMVector2LengthEst
D3DXVec2LengthSq XMVector2LengthSq
D3DXVec2Dot XMVector2Dot
D3DXVec2CCW XMVector2Cross
D3DXVec2Add XMVectorAdd
D3DXVec2Subtract XMVectorSubtract
D3DXVec2Minimize XMVectorMin
D3DXVec2Maximize XMVectorMax
D3DXVec2Scale XMVectorScale
D3DXVec2Lerp XMVectorLerp ou XMVectorLerpV
D3DXVec2Normalize XMVector2Normalize ou XMVector2NormalizeEst
D3DXVec2Hermite XMVectorHermite ou XMVectorHermiteV
D3DXVec2CatmullRom XMVectorCatmullRom ou XMVectorCatmullRomV
D3DXVec2BaryCentric XMVectorBaryCentric ou XMVectorBaryCentricV
D3DXVec2Transform XMVector2Transform
D3DXVec2TransformCoord XMVector2TransformCoord
D3DXVec2TransformNormal XMVector2TransformNormal
D3DXVec2TransformArray XMVector2TransformStream
D3DXVec2TransformCoordArray XMVector2TransformCoordStream
D3DXVec2TransformNormalArray XMVector2TransformNormalStream
D3DXVec3Length XMVector3Length ou XMVector3LengthEst
D3DXVec3LengthSq XMVector3LengthSq
D3DXVec3Dot XMVector3Dot
D3DXVec3Cross XMVector3Cross
D3DXVec3Add XMVectorAdd
D3DXVec3Subtract XMVectorSubtract
D3DXVec3Minimize XMVectorMin
D3DXVec3Maximize XMVectorMax
D3DXVec3Scale XMVectorScale
D3DXVec3Lerp XMVectorLerp ou XMVectorLerpV
D3DXVec3Normalize XMVector3Normalize ou XMVector3NormalizeEst
D3DXVec3Hermite XMVectorHermite ou XMVectorHermiteV
D3DXVec3CatmullRom XMVectorCatmullRom ou XMVectorCatmullRomV
D3DXVec3BaryCentric XMVectorBaryCentric ou XMVectorBaryCentricV
D3DXVec3Transform XMVector3Transform
D3DXVec3TransformCoord XMVector3TransformCoord
D3DXVec3TransformNormal XMVector3TransformNormal
D3DXVec3TransformArray XMVector3TransformStream
D3DXVec3TransformCoordArray XMVector3TransformCoordStream
D3DXVec3TransformNormalArray XMVector3TransformNormalStream
D3DXVec3Project XMVector3Project
D3DXVec3Unproject XMVector3Unproject
D3DXVec3ProjectArray XMVector3ProjectStream
D3DXVec3UnprojectArray XMVector3UnprojectStream
D3DXVec4Length XMVector4Length ou XMVector4LengthEst
D3DXVec4LengthSq XMVector4LengthSq
D3DXVec4Dot XMVector4Dot
D3DXVec4Add XMVectorAdd
D3DXVec4Subtract XMVectorSubtract
D3DXVec4Minimize XMVectorMin
D3DXVec4Maximize XMVectorMax
D3DXVec4Scale XMVectorScale
D3DXVec4Lerp XMVectorLerp ou XMVectorLerpV
D3DXVec4Cross XMVector4Cross
D3DXVec4Normalize XMVector4Normalize ou XMVector4NormalizeEst
D3DXVec4Hermite XMVectorHermite ou XMVectorHermiteV
D3DXVec4CatmullRom XMVectorCatmullRom ou XMVectorCatmullRomV
D3DXVec4BaryCentric XMVectorBaryCentric ou XMVectorBaryCentricV
D3DXVec4Transform XMVector4Transform
D3DXVec4TransformArray XMVector4TransformStream
D3DXMatrixIdentity XMMatrixIdentity
D3DXMatrixDeterminant XMMatrixDeterminant
D3DXMatrixDecompose XMMatrixDecompose
D3DXMatrixTranspose XMMatrixTranspose
D3DXMatrixMultiply XMMatrixMultiply
D3DXMatrixMultiplyTranspose XMMatrixMultiplyTranspose
D3DXMatrixInverse XMMatrixInverse
D3DXMatrixScaling XMMatrixScaling
D3DXMatrixTranslation XMMatrixTranslation
D3DXMatrixRotationX XMMatrixRotationX
D3DXMatrixRotationY XMMatrixRotationY
D3DXMatrixRotationZ XMMatrixRotationZ
D3DXMatrixRotationAxis XMMatrixRotationAxis
D3DXMatrixRotationQuaternion XMMatrixRotationQuaternion
D3DXMatrixRotationYawPitchRoll XMMatrixRotationRollPitchYaw (Notez que l’ordre des paramètres est différent : D3DXMatrixRotationYawPitchRoll prend yaw, pitch, roll, XMMatrixRotationRollPitchYaw prend pitch, yaw, roll)
D3DXMatrixTransformation XMMatrixTransformation
D3DXMatrixTransformation2D XMMatrixTransformation2D
D3DXMatrixAffineTransformation XMMatrixAffineTransformation
D3DXMatrixAffineTransformation2D XMMatrixAffineTransformation2D
D3DXMatrixLookAtRH XMMatrixLookAtRH
D3DXMatrixLookAtLH XMMatrixLookAtLH
D3DXMatrixPerspectiveRH XMMatrixPerspectiveRH
D3DXMatrixPerspectiveLH XMMatrixPerspectiveLH
D3DXMatrixPerspectiveFovRH XMMatrixPerspectiveFovRH
D3DXMatrixPerspectiveFovLH XMMatrixPerspectiveFovLH
D3DXMatrixPerspectiveOffCenterRH XMMatrixPerspectiveOffCenterRH
D3DXMatrixPerspectiveOffCenterLH XMMatrixPerspectiveOffCenterLH
D3DXMatrixOrthoRH XMMatrixOrthographicRH
D3DXMatrixOrthoLH XMMatrixOrthographicLH
D3DXMatrixOrthoOffCenterRH XMMatrixOrthographicOffCenterRH
D3DXMatrixOrthoOffCenterLH XMMatrixOrthographicOffCenterLH
D3DXMatrixShadow XMMatrixShadow
D3DXMatrixReflect XMMatrixReflect
D3DXQuaternionLength XMQuaternionLength
D3DXQuaternionLengthSq XMQuaternionLengthSq
D3DXQuaternionDot XMQuaternionDot
D3DXQuaternionIdentity XMQuaternionIdentity
D3DXQuaternionIsIdentity XMQuaternionIsIdentity
D3DXQuaternionConjugate XMQuaternionConjugate
D3DXQuaternionToAxisAngle XMQuaternionToAxisAngle
D3DXQuaternionRotationMatrix XMQuaternionRotationMatrix
D3DXQuaternionRotationAxis XMQuaternionRotationAxis
D3DXQuaternionRotationYawPitchRoll XMQuaternionRotationRollPitchYaw (Notez que l’ordre des paramètres est différent : D3DXQuaternionRotationYawPitchRoll prend yaw, pitch, roll, XMQuaternionRotationRollPitchYaw prend pitch, yaw, roll)
D3DXQuaternionMultiply XMQuaternionMultiply
D3DXQuaternionNormalize XMQuaternionNormalize ou XMQuaternionNormalizeEst
D3DXQuaternionInverse XMQuaternionInverse
D3DXQuaternionLn XMQuaternionLn
D3DXQuaternionExp XMQuaternionExp
D3DXQuaternionSlerp XMQuaternionSlerp ou XMQuaternionSlerpV
D3DXQuaternionSquad XMQuaternionSquad ou XMQuaternionSquadV
D3DXQuaternionSquadSetup XMQuaternionSquadSetup
D3DXQuaternionBaryCentric XMQuaternionBaryCentric ou XMQuaternionBaryCentricV
D3DXPlaneDot XMPlaneDot
D3DXPlaneDotCoord XMPlaneDotCoord
D3DXPlaneDotNormal XMPlaneDotNormal
D3DXPlaneScale XMVectorScale
D3DXPlaneNormalize XMPlaneNormalize ou XMPlaneNormalizeEst
D3DXPlaneIntersectLine XMPlaneIntersectLine
D3DXPlaneFromPointNormal XMPlaneFromPointNormal
D3DXPlaneFromPoints XMPlaneFromPoints
D3DXPlaneTransform XMPlaneTransform
D3DXPlaneTransformArray XMPlaneTransformStream
D3DXColorNegative XMColorNegative
D3DXColorAdd XMVectorAdd
D3DXColorSubtract XMVectorSubtract
D3DXColorScale XMVectorScale
D3DXColorModulate XMColorModulate
D3DXColorLerp XMVectorLerp ou XMVectorLerpV
D3DXColorAdjustSaturation XMColorAdjustSaturation
D3DXColorAdjustContrast XMColorAdjustContrast
D3DXFresnelTerm XMFresnelTerm

 

Notes

Les fonctions harmoniques sphériques pour DirectXMath sont disponibles séparément. Un DirectXMath équivalent à ID3DXMatrixStack est également disponible.

 

Guide de programmation DirectXMath