.file "atanl.s" // Copyright (C) 2000, 2001, Intel Corporation // All rights reserved. // // Contributed 2/2/2000 by John Harrison, Ted Kubaska, Bob Norin, Shane Story, // and Ping Tak Peter Tang of the Computational Software Lab, Intel Corporation. // // Redistribution and use in source and binary forms, with or without // modification, are permitted provided that the following conditions are // met: // // * Redistributions of source code must retain the above copyright // notice, this list of conditions and the following disclaimer. // // * Redistributions in binary form must reproduce the above copyright // notice, this list of conditions and the following disclaimer in the // documentation and/or other materials provided with the distribution. // // * The name of Intel Corporation may not be used to endorse or promote // products derived from this software without specific prior written // permission. // // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL OR ITS // CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, // EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, // PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR // PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY // OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY OR TORT (INCLUDING // NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS // SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. // // Intel Corporation is the author of this code, and requests that all // problem reports or change requests be submitted to it directly at // http://developer.intel.com/opensource. // // // ********************************************************************* // // History // 2/02/00 (hand-optimized) // 4/04/00 Unwind support added // 8/15/00 Bundle added after call to __libm_error_support to properly // set [the previously overwritten] GR_Parameter_RESULT. // // ********************************************************************* // // Function: atanl(x) = inverse tangent(x), for double extended x values // Function: atan2l(y,x) = atan(y/x), for double extended x values // // ********************************************************************* // // Resources Used: // // Floating-Point Registers: f8 (Input and Return Value) // f9-f15 // f32-f79 // // General Purpose Registers: // r32-r48 // r49,r50,r51,r52 (Arguments to error support for 0,0 case) // // Predicate Registers: p6-p15 // // ********************************************************************* // // IEEE Special Conditions: // // Denormal fault raised on denormal inputs // Underflow exceptions may occur // Special error handling for the y=0 and x=0 case // Inexact raised when appropriate by algorithm // // atanl(SNaN) = QNaN // atanl(QNaN) = QNaN // atanl(+/-0) = +/- 0 // atanl(+/-Inf) = +/-pi/2 // // atan2l(Any NaN for x or y) = QNaN // atan2l(+/-0,x) = +/-0 for x > 0 // atan2l(+/-0,x) = +/-pi for x < 0 // atan2l(+/-0,+0) = +/-0 // atan2l(+/-0,-0) = +/-pi // atan2l(y,+/-0) = pi/2 y > 0 // atan2l(y,+/-0) = -pi/2 y < 0 // atan2l(+/-y, Inf) = +/-0 for finite y > 0 // atan2l(+/-Inf, x) = +/-pi/2 for finite x // atan2l(+/-y, -Inf) = +/-pi for finite y > 0 // atan2l(+/-Inf, Inf) = +/-pi/4 // atan2l(+/-Inf, -Inf) = +/-3pi/4 // // ********************************************************************* // // Mathematical Description // --------------------------- // // The function ATANL( Arg_Y, Arg_X ) returns the "argument" // or the "phase" of the complex number // // Arg_X + i Arg_Y // // or equivalently, the angle in radians from the positive // x-axis to the line joining the origin and the point // (Arg_X,Arg_Y) // // // (Arg_X, Arg_Y) x // \ // \ // \ // \ // \ angle between is ATANL(Arg_Y,Arg_X) // \ // ------------------> X-axis // Origin // // Moreover, this angle is reported in the range [-pi,pi] thus // // -pi <= ATANL( Arg_Y, Arg_X ) <= pi. // // From the geometry, it is easy to define ATANL when one of // Arg_X or Arg_Y is +-0 or +-inf: // // // \ Y | // X \ | +0 | -0 | +inf | -inf | finite non-zero // \ | | | | | // ______________________________________________________ // | | | | // +-0 | Invalid/ | pi/2 | -pi/2 | sign(Y)*pi/2 // | qNaN | | | // -------------------------------------------------------- // | | | | | // +inf | +0 | -0 | pi/4 | -pi/4 | sign(Y)*0 // -------------------------------------------------------- // | | | | | // -inf | +pi | -pi | 3pi/4 | -3pi/4 | sign(Y)*pi // -------------------------------------------------------- // finite | X>0? | pi/2 | -pi/2 | normal case // non-zero| sign(Y)*0: | | | // | sign(Y)*pi | | | // // // One must take note that ATANL is NOT the arctangent of the // value Arg_Y/Arg_X; but rather ATANL and arctan are related // in a slightly more complicated way as follows: // // Let U := max(|Arg_X|, |Arg_Y|); V := min(|Arg_X|, |Arg_Y|); // sign_X be the sign bit of Arg_X, i.e., sign_X is 0 or 1; // s_X be the sign of Arg_X, i.e., s_X = (-1)^sign_X; // // sign_Y be the sign bit of Arg_Y, i.e., sign_Y is 0 or 1; // s_Y be the sign of Arg_Y, i.e., s_Y = (-1)^sign_Y; // // swap be 0 if |Arg_X| >= |Arg_Y| and 1 otherwise. // // Then, ATANL(Arg_Y, Arg_X) = // // / arctan(V/U) \ sign_X = 0 & swap = 0 // | pi/2 - arctan(V/U) | sign_X = 0 & swap = 1 // s_Y * | | // | pi - arctan(V/U) | sign_X = 1 & swap = 0 // \ pi/2 + arctan(V/U) / sign_X = 1 & swap = 1 // // // This relationship also suggest that the algorithm's major // task is to calculate arctan(V/U) for 0 < V <= U; and the // final Result is given by // // s_Y * { (P_hi + P_lo) + sigma * arctan(V/U) } // // where // // (P_hi,P_lo) represents M(sign_X,swap)*(pi/2) accurately // // M(sign_X,swap) = 0 for sign_X = 0 and swap = 0 // 1 for swap = 1 // 2 for sign_X = 1 and swap = 0 // // and // // sigma = { (sign_X XOR swap) : -1.0 : 1.0 } // // = (-1) ^ ( sign_X XOR swap ) // // Both (P_hi,P_lo) and sigma can be stored in a table and fetched // using (sign_X,swap) as an index. (P_hi, P_lo) can be stored as a // double-precision, and single-precision pair; and sigma can // obviously be just a single-precision number. // // In the algorithm we propose, arctan(V/U) is calculated to high accuracy // as A_hi + A_lo. Consequently, the Result ATANL( Arg_Y, Arg_X ) is // given by // // s_Y*P_hi + s_Y*sigma*A_hi + s_Y*(sigma*A_lo + P_lo) // // We now discuss the calculation of arctan(V/U) for 0 < V <= U. // // For (V/U) < 2^(-3), we use a simple polynomial of the form // // z + z^3*(P_1 + z^2*(P_2 + z^2*(P_3 + ... + P_8))) // // where z = V/U. // // For the sake of accuracy, the first term "z" must approximate V/U to // extra precision. For z^3 and higher power, a working precision // approximation to V/U suffices. Thus, we obtain: // // z_hi + z_lo = V/U to extra precision and // z = V/U to working precision // // The value arctan(V/U) is delivered as two pieces (A_hi, A_lo) // // (A_hi,A_lo) = (z_hi, z^3*(P_1 + ... + P_8) + z_lo). // // // For 2^(-3) <= (V/U) <= 1, we use a table-driven approach. // Consider // // (V/U) = 2^k * 1.b_1 b_2 .... b_63 b_64 b_65 .... // // Define // // z_hi = 2^k * 1.b_1 b_2 b_3 b_4 1 // // then // / \ // | (V/U) - z_hi | // arctan(V/U) = arctan(z_hi) + acrtan| -------------- | // | 1 + (V/U)*z_hi | // \ / // // / \ // | V - z_hi*U | // = arctan(z_hi) + acrtan| -------------- | // | U + V*z_hi | // \ / // // = arctan(z_hi) + acrtan( V' / U' ) // // // where // // V' = V - U*z_hi; U' = U + V*z_hi. // // Let // // w_hi + w_lo = V'/U' to extra precision and // w = V'/U' to working precision // // then we can approximate arctan(V'/U') by // // arctan(V'/U') = w_hi + w_lo // + w^3*(Q_1 + w^2*(Q_2 + w^2*(Q_3 + w^2*Q_4))) // // = w_hi + w_lo + poly // // Finally, arctan(z_hi) is calculated beforehand and stored in a table // as Tbl_hi, Tbl_lo. Thus, // // (A_hi, A_lo) = (Tbl_hi, w_hi+(poly+(w_lo+Tbl_lo))) // // This completes the mathematical description. // // // Algorithm // ------------- // // Step 0. Check for unsupported format. // // If // ( expo(Arg_X) not zero AND msb(Arg_X) = 0 ) OR // ( expo(Arg_Y) not zero AND msb(Arg_Y) = 0 ) // // then one of the arguments is unsupported. Generate an // invalid and return qNaN. // // Step 1. Initialize // // Normalize Arg_X and Arg_Y and set the following // // sign_X := sign_bit(Arg_X) // s_Y := (sign_bit(Arg_Y)==0? 1.0 : -1.0) // swap := (|Arg_X| >= |Arg_Y|? 0 : 1 ) // U := max( |Arg_X|, |Arg_Y| ) // V := min( |Arg_X|, |Arg_Y| ) // // execute: frcap E, pred, V, U // If pred is 0, go to Step 5 for special cases handling. // // Step 2. Decide on branch. // // Q := E * V // If Q < 2^(-3) go to Step 4 for simple polynomial case. // // Step 3. Table-driven algorithm. // // Q is represented as // // 2^(-k) * 1.b_1 b_2 b_3 ... b_63; k = 0,-1,-2,-3 // // and that if k = 0, b_1 = b_2 = b_3 = b_4 = 0. // // Define // // z_hi := 2^(-k) * 1.b_1 b_2 b_3 b_4 1 // // (note that there are 49 possible values of z_hi). // // ...We now calculate V' and U'. While V' is representable // ...as a 64-bit number because of cancellation, U' is // ...not in general a 64-bit number. Obtaining U' accurately // ...requires two working precision numbers // // U_prime_hi := U + V * z_hi ...WP approx. to U' // U_prime_lo := ( U - U_prime_hi ) + V*z_hi ...observe order // V_prime := V - U * z_hi ...this is exact // // C_hi := frcpa (1.0, U_prime_hi) ...C_hi approx 1/U'_hi // // loop 3 times // C_hi := C_hi + C_hi*(1.0 - C_hi*U_prime_hi) // // ...at this point C_hi is (1/U_prime_hi) to roughly 64 bits // // w_hi := V_prime * C_hi ...w_hi is V_prime/U_prime to // ...roughly working precision // // ...note that we want w_hi + w_lo to approximate // ...V_prime/(U_prime_hi + U_prime_lo) to extra precision // ...but for now, w_hi is good enough for the polynomial // ...calculation. // // wsq := w_hi*w_hi // poly := w_hi*wsq*(Q_1 + wsq*(Q_2 + wsq*(Q_3 + wsq*Q_4))) // // Fetch // (Tbl_hi, Tbl_lo) = atan(z_hi) indexed by (k,b_1,b_2,b_3,b_4) // ...Tbl_hi is a double-precision number // ...Tbl_lo is a single-precision number // // (P_hi, P_lo) := M(sign_X,swap)*(Pi_by_2_hi, Pi_by_2_lo) // ...as discussed previous. Again; the implementation can // ...chose to fetch P_hi and P_lo from a table indexed by // ...(sign_X, swap). // ...P_hi is a double-precision number; // ...P_lo is a single-precision number. // // ...calculate w_lo so that w_hi + w_lo is V'/U' accurately // w_lo := ((V_prime - w_hi*U_prime_hi) - // w_hi*U_prime_lo) * C_hi ...observe order // // // ...Ready to deliver arctan(V'/U') as A_hi, A_lo // A_hi := Tbl_hi // A_lo := w_hi + (poly + (Tbl_lo + w_lo)) ...observe order // // ...Deliver final Result // ...s_Y*P_hi + s_Y*sigma*A_hi + s_Y*(sigma*A_lo + P_lo) // // sigma := ( (sign_X XOR swap) ? -1.0 : 1.0 ) // ...sigma can be obtained by a table lookup using // ...(sign_X,swap) as index and stored as single precision // ...sigma should be calculated earlier // // P_hi := s_Y*P_hi // A_hi := s_Y*A_hi // // Res_hi := P_hi + sigma*A_hi ...this is exact because // ...both P_hi and Tbl_hi // ...are double-precision // ...and |Tbl_hi| > 2^(-4) // ...P_hi is either 0 or // ...between (1,4) // // Res_lo := sigma*A_lo + P_lo // // Return Res_hi + s_Y*Res_lo in user-defined rounding control // // Step 4. Simple polynomial case. // // ...E and Q are inherited from Step 2. // // A_hi := Q ...Q is inherited from Step 2 Q approx V/U // // loop 3 times // E := E + E2(1.0 - E*U1 // ...at this point E approximates 1/U to roughly working precision // // z := V * E ...z approximates V/U to roughly working precision // zsq := z * z // z8 := zsq * zsq; z8 := z8 * z8 // // poly1 := P_4 + zsq*(P_5 + zsq*(P_6 + zsq*(P_7 + zsq*P_8))) // poly2 := zsq*(P_1 + zsq*(P_2 + zsq*P_3)) // // poly := poly1 + z8*poly2 // // z_lo := (V - A_hi*U)*E // // A_lo := z*poly + z_lo // ...A_hi, A_lo approximate arctan(V/U) accurately // // (P_hi, P_lo) := M(sign_X,swap)*(Pi_by_2_hi, Pi_by_2_lo) // ...one can store the M(sign_X,swap) as single precision // ...values // // ...Deliver final Result // ...s_Y*P_hi + s_Y*sigma*A_hi + s_Y*(sigma*A_lo + P_lo) // // sigma := ( (sign_X XOR swap) ? -1.0 : 1.0 ) // ...sigma can be obtained by a table lookup using // ...(sign_X,swap) as index and stored as single precision // ...sigma should be calculated earlier // // P_hi := s_Y*P_hi // A_hi := s_Y*A_hi // // Res_hi := P_hi + sigma*A_hi ...need to compute // ...P_hi + sigma*A_hi // ...exactly // // tmp := (P_hi - Res_hi) + sigma*A_hi // // Res_lo := s_Y*(sigma*A_lo + P_lo) + tmp // // Return Res_hi + Res_lo in user-defined rounding control // // Step 5. Special Cases // // If pred is 0 where pred is obtained in // frcap E, pred, V, U // // we are in one of those special cases of 0,+-inf or NaN // // If one of U and V is NaN, return U+V (which will generate // invalid in case one is a signaling NaN). Otherwise, // return the Result as described in the table // // // // \ Y | // X \ | +0 | -0 | +inf | -inf | finite non-zero // \ | | | | | // ______________________________________________________ // | | | | // +-0 | Invalid/ | pi/2 | -pi/2 | sign(Y)*pi/2 // | qNaN | | | // -------------------------------------------------------- // | | | | | // +inf | +0 | -0 | pi/4 | -pi/4 | sign(Y)*0 // -------------------------------------------------------- // | | | | | // -inf | +pi | -pi | 3pi/4 | -3pi/4 | sign(Y)*pi // -------------------------------------------------------- // finite | X>0? | pi/2 | -pi/2 | // non-zero| sign(Y)*0: | | | N/A // | sign(Y)*pi | | | // // #include "libm_support.h" ArgY_orig = f8 Result = f8 FR_RESULT = f8 ArgX_orig = f9 ArgX = f10 FR_X = f10 ArgY = f11 FR_Y = f11 s_Y = f12 U = f13 V = f14 E = f15 Q = f32 z_hi = f33 U_prime_hi = f34 U_prime_lo = f35 V_prime = f36 C_hi = f37 w_hi = f38 w_lo = f39 wsq = f40 poly = f41 Tbl_hi = f42 Tbl_lo = f43 P_hi = f44 P_lo = f45 A_hi = f46 A_lo = f47 sigma = f48 Res_hi = f49 Res_lo = f50 Z = f52 zsq = f53 z8 = f54 poly1 = f55 poly2 = f56 z_lo = f57 tmp = f58 P_1 = f59 Q_1 = f60 P_2 = f61 Q_2 = f62 P_3 = f63 Q_3 = f64 P_4 = f65 Q_4 = f66 P_5 = f67 P_6 = f68 P_7 = f69 P_8 = f70 TWO_TO_NEG3 = f71 U_hold = f72 C_hi_hold = f73 E_hold = f74 M = f75 ArgX_abs = f76 ArgY_abs = f77 Result_lo = f78 A_temp = f79 GR_SAVE_PFS = r33 GR_SAVE_B0 = r34 GR_SAVE_GP = r35 sign_X = r36 sign_Y = r37 swap = r38 table_ptr1 = r39 table_ptr2 = r40 k = r41 lookup = r42 exp_ArgX = r43 exp_ArgY = r44 exponent_Q = r45 significand_Q = r46 special = r47 special1 = r48 GR_Parameter_X = r49 GR_Parameter_Y = r50 GR_Parameter_RESULT = r51 GR_Parameter_TAG = r52 int_temp = r52 #ifdef _LIBC .rodata #else .data #endif .align 64 Constants_atan: ASM_TYPE_DIRECTIVE(Constants_atan,@object) data4 0x54442D18, 0x3FF921FB, 0x248D3132, 0x3E000000 // double pi/2, single lo_pi/2, two**(-3) data4 0xAAAAAAA3, 0xAAAAAAAA, 0x0000BFFD, 0x00000000 // P_1 data4 0xCCCC54B2, 0xCCCCCCCC, 0x00003FFC, 0x00000000 // P_2 data4 0x47E4D0C2, 0x92492492, 0x0000BFFC, 0x00000000 // P_3 data4 0x58870889, 0xE38E38E0, 0x00003FFB, 0x00000000 // P_4 data4 0x290149F8, 0xBA2E895B, 0x0000BFFB, 0x00000000 // P_5 data4 0x250F733D, 0x9D88E6D4, 0x00003FFB, 0x00000000 // P_6 data4 0xFB8745A0, 0x884E51FF, 0x0000BFFB, 0x00000000 // P_7 data4 0x394396BD, 0xE1C7412B, 0x00003FFA, 0x00000000 // P_8 data4 0xAAAAA52F, 0xAAAAAAAA, 0x0000BFFD, 0x00000000 // Q_1 data4 0xC75B60D3, 0xCCCCCCCC, 0x00003FFC, 0x00000000 // Q_2 data4 0x011F1940, 0x924923AD, 0x0000BFFC, 0x00000000 // Q_3 data4 0x2A5F89BD, 0xE36F716D, 0x00003FFB, 0x00000000 // Q_4 // // Entries Tbl_hi (double precision) // B = 1+Index/16+1/32 Index = 0 // Entries Tbl_lo (single precision) // B = 1+Index/16+1/32 Index = 0 // data4 0xA935BD8E, 0x3FE9A000, 0x23ACA08F, 0x00000000 // // Entries Tbl_hi (double precision) Index = 0,1,...,15 // B = 2^(-1)*(1+Index/16+1/32) // Entries Tbl_lo (single precision) // Index = 0,1,...,15 B = 2^(-1)*(1+Index/16+1/32) // data4 0x7F175A34, 0x3FDE77EB, 0x238729EE, 0x00000000 data4 0x73C1A40B, 0x3FE0039C, 0x249334DB, 0x00000000 data4 0x5B5B43DA, 0x3FE0C614, 0x22CBA7D1, 0x00000000 data4 0x88BE7C13, 0x3FE1835A, 0x246310E7, 0x00000000 data4 0xE2CC9E6A, 0x3FE23B71, 0x236210E5, 0x00000000 data4 0x8406CBCA, 0x3FE2EE62, 0x2462EAF5, 0x00000000 data4 0x1CD41719, 0x3FE39C39, 0x24B73EF3, 0x00000000 data4 0x5B795B55, 0x3FE44506, 0x24C11260, 0x00000000 data4 0x5BB6EC04, 0x3FE4E8DE, 0x242519EE, 0x00000000 data4 0x1F732FBA, 0x3FE587D8, 0x24D4346C, 0x00000000 data4 0x115D7B8D, 0x3FE6220D, 0x24ED487B, 0x00000000 data4 0x920B3D98, 0x3FE6B798, 0x2495FF1E, 0x00000000 data4 0x8FBA8E0F, 0x3FE74897, 0x223D9531, 0x00000000 data4 0x289FA093, 0x3FE7D528, 0x242B0411, 0x00000000 data4 0x576CC2C5, 0x3FE85D69, 0x2335B374, 0x00000000 data4 0xA99CC05D, 0x3FE8E17A, 0x24C27CFB, 0x00000000 // // Entries Tbl_hi (double precision) Index = 0,1,...,15 // B = 2^(-2)*(1+Index/16+1/32) // Entries Tbl_lo (single precision) // Index = 0,1,...,15 B = 2^(-2)*(1+Index/16+1/32) // data4 0x510665B5, 0x3FD025FA, 0x24263482, 0x00000000 data4 0x362431C9, 0x3FD1151A, 0x242C8DC9, 0x00000000 data4 0x67E47C95, 0x3FD20255, 0x245CF9BA, 0x00000000 data4 0x7A823CFE, 0x3FD2ED98, 0x235C892C, 0x00000000 data4 0x29271134, 0x3FD3D6D1, 0x2389BE52, 0x00000000 data4 0x586890E6, 0x3FD4BDEE, 0x24436471, 0x00000000 data4 0x175E0F4E, 0x3FD5A2E0, 0x2389DBD4, 0x00000000 data4 0x9F5FA6FD, 0x3FD68597, 0x2476D43F, 0x00000000 data4 0x52817501, 0x3FD76607, 0x24711774, 0x00000000 data4 0xB8DF95D7, 0x3FD84422, 0x23EBB501, 0x00000000 data4 0x7CD0C662, 0x3FD91FDE, 0x23883A0C, 0x00000000 data4 0x66168001, 0x3FD9F930, 0x240DF63F, 0x00000000 data4 0x5422058B, 0x3FDAD00F, 0x23FE261A, 0x00000000 data4 0x378624A5, 0x3FDBA473, 0x23A8CD0E, 0x00000000 data4 0x0AAD71F8, 0x3FDC7655, 0x2422D1D0, 0x00000000 data4 0xC9EC862B, 0x3FDD45AE, 0x2344A109, 0x00000000 // // Entries Tbl_hi (double precision) Index = 0,1,...,15 // B = 2^(-3)*(1+Index/16+1/32) // Entries Tbl_lo (single precision) // Index = 0,1,...,15 B = 2^(-3)*(1+Index/16+1/32) // data4 0x84212B3D, 0x3FC068D5, 0x239874B6, 0x00000000 data4 0x41060850, 0x3FC16465, 0x2335E774, 0x00000000 data4 0x171A535C, 0x3FC25F6E, 0x233E36BE, 0x00000000 data4 0xEDEB99A3, 0x3FC359E8, 0x239680A3, 0x00000000 data4 0xC6092A9E, 0x3FC453CE, 0x230FB29E, 0x00000000 data4 0xBA11570A, 0x3FC54D18, 0x230C1418, 0x00000000 data4 0xFFB3AA73, 0x3FC645BF, 0x23F0564A, 0x00000000 data4 0xE8A7D201, 0x3FC73DBD, 0x23D4A5E1, 0x00000000 data4 0xE398EBC7, 0x3FC8350B, 0x23D4ADDA, 0x00000000 data4 0x7D050271, 0x3FC92BA3, 0x23BCB085, 0x00000000 data4 0x601081A5, 0x3FCA217E, 0x23BC841D, 0x00000000 data4 0x574D780B, 0x3FCB1696, 0x23CF4A8E, 0x00000000 data4 0x4D768466, 0x3FCC0AE5, 0x23BECC90, 0x00000000 data4 0x4E1D5395, 0x3FCCFE65, 0x2323DCD2, 0x00000000 data4 0x864C9D9D, 0x3FCDF110, 0x23F53F3A, 0x00000000 data4 0x451D980C, 0x3FCEE2E1, 0x23CCB11F, 0x00000000 data4 0x54442D18, 0x400921FB, 0x33145C07, 0x3CA1A626 // PI two doubles data4 0x54442D18, 0x3FF921FB, 0x33145C07, 0x3C91A626 // PI_by_2 two dbles data4 0x54442D18, 0x3FE921FB, 0x33145C07, 0x3C81A626 // PI_by_4 two dbles data4 0x7F3321D2, 0x4002D97C, 0x4C9E8A0A, 0x3C9A7939 // 3PI_by_4 two dbles ASM_SIZE_DIRECTIVE(Constants_atan) .text .proc atanl# .global atanl# .align 64 atanl: { .mfb nop.m 999 (p0) mov ArgX_orig = f1 (p0) br.cond.sptk atan2l ;; } .endp atanl ASM_SIZE_DIRECTIVE(atanl) .text .proc atan2l# .global atan2l# #ifdef _LIBC .proc __atan2l# .global __atan2l# .proc __ieee754_atan2l# .global __ieee754_atan2l# #endif .align 64 atan2l: #ifdef _LIBC __atan2l: __ieee754_atan2l: #endif { .mfi alloc r32 = ar.pfs, 0, 17 , 4, 0 (p0) mov ArgY = ArgY_orig } { .mfi nop.m 999 (p0) mov ArgX = ArgX_orig nop.i 999 };; { .mfi nop.m 999 (p0) fclass.m.unc p7,p0 = ArgY_orig, 0x103 nop.i 999 } { .mfi nop.m 999 // // // Save original input args and load table ptr. // (p0) fclass.m.unc p6,p0 = ArgX_orig, 0x103 nop.i 999 };; { .mfi (p0) addl table_ptr1 = @ltoff(Constants_atan#), gp (p0) fclass.m.unc p0,p9 = ArgY_orig, 0x1FF nop.i 999 ;; } { .mfi ld8 table_ptr1 = [table_ptr1] (p0) fclass.m.unc p0,p8 = ArgX_orig, 0x1FF nop.i 999 } { .mfi nop.m 999 (p0) fclass.m.unc p13,p0 = ArgY_orig, 0x0C3 nop.i 999 ;; } { .mfi (p0) fclass.m.unc p12,p0 = ArgX_orig, 0x0C3 nop.i 999 } // // Check for NatVals. // Check for everything - if false, then must be pseudo-zero // or pseudo-nan (IA unsupporteds). // { .mib nop.m 999 nop.i 999 (p6) br.cond.spnt L(ATANL_NATVAL) ;; } { .mib nop.m 999 nop.i 999 (p7) br.cond.spnt L(ATANL_NATVAL) ;; } { .mib (p0) ldfd P_hi = [table_ptr1],8 nop.i 999 (p8) br.cond.spnt L(ATANL_UNSUPPORTED) ;; } { .mbb (p0) add table_ptr2 = 96, table_ptr1 (p9) br.cond.spnt L(ATANL_UNSUPPORTED) // // Load double precision high-order part of pi // (p12) br.cond.spnt L(ATANL_NAN) ;; } { .mfb nop.m 999 (p0) fnorm.s1 ArgX = ArgX (p13) br.cond.spnt L(ATANL_NAN) ;; } // // Normalize the input argument. // Branch out if NaN inputs // { .mmf (p0) ldfs P_lo = [table_ptr1], 4 nop.m 999 (p0) fnorm.s1 ArgY = ArgY ;; } { .mmf nop.m 999 (p0) ldfs TWO_TO_NEG3 = [table_ptr1], 180 // // U = max(ArgX_abs,ArgY_abs) // V = min(ArgX_abs,ArgY_abs) // if PR1, swap = 0 // if PR2, swap = 1 // (p0) mov M = f1 ;; } { .mfi nop.m 999 // // Get exp and sign of ArgX // Get exp and sign of ArgY // Load 2**(-3) and increment ptr to Q_4. // (p0) fmerge.s ArgX_abs = f1, ArgX nop.i 999 ;; } // // load single precision low-order part of pi = P_lo // { .mfi (p0) getf.exp sign_X = ArgX (p0) fmerge.s ArgY_abs = f1, ArgY nop.i 999 ;; } { .mii (p0) getf.exp sign_Y = ArgY nop.i 999 ;; (p0) shr sign_X = sign_X, 17 ;; } { .mii nop.m 999 (p0) shr sign_Y = sign_Y, 17 ;; (p0) cmp.eq.unc p8, p9 = 0x00000, sign_Y ;; } { .mfi nop.m 999 // // Is ArgX_abs >= ArgY_abs // Is sign_Y == 0? // (p0) fmax.s1 U = ArgX_abs, ArgY_abs nop.i 999 } { .mfi nop.m 999 // // ArgX_abs = |ArgX| // ArgY_abs = |ArgY| // sign_X is sign bit of ArgX // sign_Y is sign bit of ArgY // (p0) fcmp.ge.s1 p6, p7 = ArgX_abs, ArgY_abs nop.i 999 ;; } { .mfi nop.m 999 (p0) fmin.s1 V = ArgX_abs, ArgY_abs nop.i 999 ;; } { .mfi nop.m 999 (p8) fadd.s1 s_Y = f0, f1 (p6) cmp.eq.unc p10, p11 = 0x00000, sign_X } { .mii (p6) add swap = r0, r0 nop.i 999 ;; (p7) add swap = 1, r0 } { .mfi nop.m 999 // // Let M = 1.0 // if p8, s_Y = 1.0 // if p9, s_Y = -1.0 // (p10) fsub.s1 M = M, f1 nop.i 999 ;; } { .mfi nop.m 999 (p9) fsub.s1 s_Y = f0, f1 nop.i 999 ;; } { .mfi nop.m 999 (p0) frcpa.s1 E, p6 = V, U nop.i 999 ;; } { .mbb nop.m 999 // // E = frcpa(V,U) // (p6) br.cond.sptk L(ATANL_STEP2) (p0) br.cond.spnt L(ATANL_SPECIAL_HANDLING) ;; } L(ATANL_STEP2): { .mfi nop.m 999 (p0) fmpy.s1 Q = E, V nop.i 999 } { .mfi nop.m 999 (p0) fcmp.eq.s0 p0, p9 = f1, ArgY_orig nop.i 999 ;; } { .mfi nop.m 999 // // Is Q < 2**(-3)? // (p0) fcmp.eq.s0 p0, p8 = f1, ArgX_orig nop.i 999 } { .mfi nop.m 999 (p11) fadd.s1 M = M, f1 nop.i 999 ;; } { .mlx nop.m 999 // ************************************************* // ********************* STEP2 ********************* // ************************************************* (p0) movl special = 0x8400000000000000 } { .mlx nop.m 999 // // lookup = b_1 b_2 b_3 B_4 // (p0) movl special1 = 0x0000000000000100 ;; } { .mfi nop.m 999 // // Do fnorms to raise any denormal operand // exceptions. // (p0) fmpy.s1 P_hi = M, P_hi nop.i 999 } { .mfi nop.m 999 (p0) fmpy.s1 P_lo = M, P_lo nop.i 999 ;; } { .mfi nop.m 999 // // Q = E * V // (p0) fcmp.lt.unc.s1 p6, p7 = Q, TWO_TO_NEG3 nop.i 999 ;; } { .mmb (p0) getf.sig significand_Q = Q (p0) getf.exp exponent_Q = Q nop.b 999 ;; } { .mmi nop.m 999 ;; (p0) andcm k = 0x0003, exponent_Q (p0) extr.u lookup = significand_Q, 59, 4 ;; } { .mib nop.m 999 (p0) dep special = lookup, special, 59, 4 // // Generate 1.b_1 b_2 b_3 b_4 1 0 0 0 ... 0 // (p6) br.cond.spnt L(ATANL_POLY) ;; } { .mfi (p0) cmp.eq.unc p8, p9 = 0x0000, k (p0) fmpy.s1 P_hi = s_Y, P_hi // // We waited a few extra cycles so P_lo and P_hi could be calculated. // Load the constant 256 for loading up table entries. // // ************************************************* // ******************** STEP3 ********************** // ************************************************* (p0) add table_ptr2 = 16, table_ptr1 } // // Let z_hi have exponent and sign of original Q // Load the Tbl_hi(0) else, increment pointer. // { .mii (p0) ldfe Q_4 = [table_ptr1], -16 (p0) xor swap = sign_X, swap ;; (p9) sub k = k, r0, 1 } { .mmi (p0) setf.sig z_hi = special (p0) ldfe Q_3 = [table_ptr1], -16 (p9) add table_ptr2 = 16, table_ptr2 ;; } // // U_hold = U - U_prime_hi // k = k * 256 - Result can be 0, 256, or 512. // { .mmb (p0) ldfe Q_2 = [table_ptr1], -16 (p8) ldfd Tbl_hi = [table_ptr2], 8 nop.b 999 ;; } // // U_prime_lo = U_hold + V * z_hi // lookup -> lookup * 16 + k // { .mmi (p0) ldfe Q_1 = [table_ptr1], -16 ;; (p8) ldfs Tbl_lo = [table_ptr2], 8 // // U_prime_hi = U + V * z_hi // Load the Tbl_lo(0) // (p9) pmpy2.r k = k, special1 ;; } { .mii nop.m 999 nop.i 999 nop.i 999 ;; } { .mii nop.m 999 nop.i 999 nop.i 999 ;; } { .mii nop.m 999 nop.i 999 nop.i 999 ;; } { .mii nop.m 999 nop.i 999 ;; (p9) shladd lookup = lookup, 0x0004, k ;; } { .mmi (p9) add table_ptr2 = table_ptr2, lookup ;; // // V_prime = V - U * z_hi // (p9) ldfd Tbl_hi = [table_ptr2], 8 nop.i 999 ;; } { .mmf nop.m 999 // // C_hi = frcpa(1,U_prime_hi) // (p9) ldfs Tbl_lo = [table_ptr2], 8 // // z_hi = s exp 1.b_1 b_2 b_3 b_4 1 0 0 0 ... 0 // Point to beginning of Tbl_hi entries - k = 0. // (p0) fmerge.se z_hi = Q, z_hi ;; } { .mfi nop.m 999 (p0) fma.s1 U_prime_hi = V, z_hi, U nop.i 999 } { .mfi nop.m 999 (p0) fnma.s1 V_prime = U, z_hi, V nop.i 999 ;; } { .mfi nop.m 999 (p0) mov A_hi = Tbl_hi nop.i 999 ;; } { .mfi nop.m 999 (p0) fsub.s1 U_hold = U, U_prime_hi nop.i 999 ;; } { .mfi nop.m 999 (p0) frcpa.s1 C_hi, p6 = f1, U_prime_hi nop.i 999 ;; } { .mfi (p0) cmp.eq.unc p7, p6 = 0x00000, swap (p0) fmpy.s1 A_hi = s_Y, A_hi nop.i 999 ;; } { .mfi nop.m 999 // // poly = wsq * poly // (p7) fadd.s1 sigma = f0, f1 nop.i 999 ;; } { .mfi nop.m 999 (p0) fma.s1 U_prime_lo = z_hi, V, U_hold nop.i 999 } { .mfi nop.m 999 (p6) fsub.s1 sigma = f0, f1 nop.i 999 ;; } { .mfi nop.m 999 (p0) fnma.s1 C_hi_hold = C_hi, U_prime_hi, f1 nop.i 999 ;; } { .mfi nop.m 999 // // A_lo = A_lo + w_hi // A_hi = s_Y * A_hi // (p0) fma.s1 Res_hi = sigma, A_hi, P_hi nop.i 999 ;; } { .mfi nop.m 999 // // C_hi_hold = 1 - C_hi * U_prime_hi (1) // (p0) fma.s1 C_hi = C_hi_hold, C_hi, C_hi nop.i 999 ;; } { .mfi nop.m 999 // // C_hi = C_hi + C_hi * C_hi_hold (1) // (p0) fnma.s1 C_hi_hold = C_hi, U_prime_hi, f1 nop.i 999 ;; } { .mfi nop.m 999 // // C_hi_hold = 1 - C_hi * U_prime_hi (2) // (p0) fma.s1 C_hi = C_hi_hold, C_hi, C_hi nop.i 999 ;; } { .mfi nop.m 999 // // C_hi = C_hi + C_hi * C_hi_hold (2) // (p0) fnma.s1 C_hi_hold = C_hi, U_prime_hi, f1 nop.i 999 ;; } { .mfi nop.m 999 // // C_hi_hold = 1 - C_hi * U_prime_hi (3) // (p0) fma.s1 C_hi = C_hi_hold, C_hi, C_hi nop.i 999 ;; } { .mfi nop.m 999 // // C_hi = C_hi + C_hi * C_hi_hold (3) // (p0) fmpy.s1 w_hi = V_prime, C_hi nop.i 999 ;; } { .mfi nop.m 999 // // w_hi = V_prime * C_hi // (p0) fmpy.s1 wsq = w_hi, w_hi nop.i 999 } { .mfi nop.m 999 (p0) fnma.s1 w_lo = w_hi, U_prime_hi, V_prime nop.i 999 ;; } { .mfi nop.m 999 // // wsq = w_hi * w_hi // w_lo = = V_prime - w_hi * U_prime_hi // (p0) fma.s1 poly = wsq, Q_4, Q_3 nop.i 999 } { .mfi nop.m 999 (p0) fnma.s1 w_lo = w_hi, U_prime_lo, w_lo nop.i 999 ;; } { .mfi nop.m 999 // // poly = Q_3 + wsq * Q_4 // w_lo = = w_lo - w_hi * U_prime_lo // (p0) fma.s1 poly = wsq, poly, Q_2 nop.i 999 } { .mfi nop.m 999 (p0) fmpy.s1 w_lo = C_hi, w_lo nop.i 999 ;; } { .mfi nop.m 999 // // poly = Q_2 + wsq * poly // w_lo = = w_lo * C_hi // (p0) fma.s1 poly = wsq, poly, Q_1 nop.i 999 } { .mfi nop.m 999 (p0) fadd.s1 A_lo = Tbl_lo, w_lo nop.i 999 ;; } { .mfi nop.m 999 // // Result = Res_hi + Res_lo * s_Y (User Supplied Rounding Mode) // (p0) fmpy.s0 Q_1 = Q_1, Q_1 nop.i 999 ;; } { .mfi nop.m 999 // // poly = Q_1 + wsq * poly // A_lo = Tbl_lo + w_lo // swap = xor(swap,sign_X) // (p0) fmpy.s1 poly = wsq, poly nop.i 999 ;; } { .mfi nop.m 999 // // Is (swap) != 0 ? // poly = wsq * poly // A_hi = Tbl_hi // (p0) fmpy.s1 poly = w_hi, poly nop.i 999 ;; } { .mfi nop.m 999 // // if (PR_1) sigma = -1.0 // if (PR_2) sigma = 1.0 // (p0) fadd.s1 A_lo = A_lo, poly nop.i 999 ;; } { .mfi nop.m 999 // // P_hi = s_Y * P_hi // A_lo = A_lo + poly // (p0) fadd.s1 A_lo = A_lo, w_hi nop.i 999 ;; } { .mfi nop.m 999 (p0) fma.s1 Res_lo = sigma, A_lo, P_lo nop.i 999 ;; } { .mfb nop.m 999 // // Res_hi = P_hi + sigma * A_hi // Res_lo = P_lo + sigma * A_lo // (p0) fma.s0 Result = Res_lo, s_Y, Res_hi // // Raise inexact. // br.ret.sptk b0 ;; } // // poly1 = P_5 + zsq * poly1 // poly2 = zsq * poly2 // L(ATANL_POLY): { .mmf (p0) xor swap = sign_X, swap nop.m 999 (p0) fnma.s1 E_hold = E, U, f1 ;; } { .mfi nop.m 999 (p0) mov A_temp = Q // // poly1 = P_4 + zsq * poly1 // swap = xor(swap,sign_X) // // sign_X gr_002 // swap gr_004 // poly1 = poly1 <== Done with poly1 // poly1 = P_4 + zsq * poly1 // swap = xor(swap,sign_X) // (p0) cmp.eq.unc p7, p6 = 0x00000, swap } { .mfi nop.m 999 (p0) fmpy.s1 P_hi = s_Y, P_hi nop.i 999 ;; } { .mfi nop.m 999 (p6) fsub.s1 sigma = f0, f1 nop.i 999 } { .mfi nop.m 999 (p7) fadd.s1 sigma = f0, f1 nop.i 999 ;; } // *********************************************** // ******************** STEP4 ******************** // *********************************************** { .mmi nop.m 999 (p0) addl table_ptr1 = @ltoff(Constants_atan#), gp nop.i 999 } ;; { .mmi ld8 table_ptr1 = [table_ptr1] nop.m 999 nop.i 999 } ;; { .mfi nop.m 999 (p0) fma.s1 E = E, E_hold, E // // Following: // Iterate 3 times E = E + E*(1.0 - E*U) // Also load P_8, P_7, P_6, P_5, P_4 // E_hold = 1.0 - E * U (1) // A_temp = Q // (p0) add table_ptr1 = 128, table_ptr1 ;; } { .mmf nop.m 999 // // E = E + E_hold*E (1) // Point to P_8. // (p0) ldfe P_8 = [table_ptr1], -16 // // poly = z8*poly1 + poly2 (Typo in writeup) // Is (swap) != 0 ? // (p0) fnma.s1 z_lo = A_temp, U, V ;; } { .mmb nop.m 999 // // E_hold = 1.0 - E * U (2) // (p0) ldfe P_7 = [table_ptr1], -16 nop.b 999 ;; } { .mmb nop.m 999 // // E = E + E_hold*E (2) // (p0) ldfe P_6 = [table_ptr1], -16 nop.b 999 ;; } { .mmb nop.m 999 // // E_hold = 1.0 - E * U (3) // (p0) ldfe P_5 = [table_ptr1], -16 nop.b 999 ;; } { .mmf nop.m 999 // // E = E + E_hold*E (3) // // // At this point E approximates 1/U to roughly working precision // z = V*E approximates V/U // (p0) ldfe P_4 = [table_ptr1], -16 (p0) fnma.s1 E_hold = E, U, f1 ;; } { .mmb nop.m 999 // // Z = V * E // (p0) ldfe P_3 = [table_ptr1], -16 nop.b 999 ;; } { .mmb nop.m 999 // // zsq = Z * Z // (p0) ldfe P_2 = [table_ptr1], -16 nop.b 999 ;; } { .mmb nop.m 999 // // z8 = zsq * zsq // (p0) ldfe P_1 = [table_ptr1], -16 nop.b 999 ;; } { .mlx nop.m 999 (p0) movl int_temp = 0x24005 } { .mfi nop.m 999 (p0) fma.s1 E = E, E_hold, E nop.i 999 ;; } { .mfi nop.m 999 (p0) fnma.s1 E_hold = E, U, f1 nop.i 999 ;; } { .mfi nop.m 999 (p0) fma.s1 E = E, E_hold, E nop.i 999 ;; } { .mfi nop.m 999 (p0) fmpy.s1 Z = V, E nop.i 999 } { .mfi nop.m 999 // // z_lo = V - A_temp * U // if (PR_2) sigma = 1.0 // (p0) fmpy.s1 z_lo = z_lo, E nop.i 999 ;; } { .mfi nop.m 999 (p0) fmpy.s1 zsq = Z, Z nop.i 999 } { .mfi nop.m 999 // // z_lo = z_lo * E // if (PR_1) sigma = -1.0 // (p0) fadd.s1 A_hi = A_temp, z_lo nop.i 999 ;; } { .mfi nop.m 999 // // z8 = z8 * z8 // // // Now what we want to do is // poly1 = P_4 + zsq*(P_5 + zsq*(P_6 + zsq*(P_7 + zsq*P_8))) // poly2 = zsq*(P_1 + zsq*(P_2 + zsq*P_3)) // (p0) fma.s1 poly1 = zsq, P_8, P_7 nop.i 999 } { .mfi nop.m 999 (p0) fma.s1 poly2 = zsq, P_3, P_2 nop.i 999 ;; } { .mfi nop.m 999 (p0) fmpy.s1 z8 = zsq, zsq nop.i 999 } { .mfi nop.m 999 (p0) fsub.s1 A_temp = A_temp, A_hi nop.i 999 ;; } { .mfi nop.m 999 // // A_lo = Z * poly + z_lo // (p0) fmerge.s tmp = A_hi, A_hi nop.i 999 ;; } { .mfi nop.m 999 // // poly1 = P_7 + zsq * P_8 // poly2 = P_2 + zsq * P_3 // (p0) fma.s1 poly1 = zsq, poly1, P_6 nop.i 999 } { .mfi nop.m 999 (p0) fma.s1 poly2 = zsq, poly2, P_1 nop.i 999 ;; } { .mfi nop.m 999 (p0) fmpy.s1 z8 = z8, z8 nop.i 999 } { .mfi nop.m 999 (p0) fadd.s1 z_lo = A_temp, z_lo nop.i 999 ;; } { .mfi nop.m 999 // // poly1 = P_6 + zsq * poly1 // poly2 = P_2 + zsq * poly2 // (p0) fma.s1 poly1 = zsq, poly1, P_5 nop.i 999 } { .mfi nop.m 999 (p0) fmpy.s1 poly2 = poly2, zsq nop.i 999 ;; } { .mfi nop.m 999 // // Result = Res_hi + Res_lo (User Supplied Rounding Mode) // (p0) fmpy.s1 P_5 = P_5, P_5 nop.i 999 ;; } { .mfi nop.m 999 (p0) fma.s1 poly1 = zsq, poly1, P_4 nop.i 999 ;; } { .mfi nop.m 999 (p0) fma.s1 poly = z8, poly1, poly2 nop.i 999 ;; } { .mfi nop.m 999 // // Fixup added to force inexact later - // A_hi = A_temp + z_lo // z_lo = (A_temp - A_hi) + z_lo // (p0) fma.s1 A_lo = Z, poly, z_lo nop.i 999 ;; } { .mfi nop.m 999 (p0) fadd.s1 A_hi = tmp, A_lo nop.i 999 ;; } { .mfi nop.m 999 (p0) fsub.s1 tmp = tmp, A_hi nop.i 999 } { .mfi nop.m 999 (p0) fmpy.s1 A_hi = s_Y, A_hi nop.i 999 ;; } { .mfi nop.m 999 (p0) fadd.s1 A_lo = tmp, A_lo nop.i 999 } { .mfi (p0) setf.exp tmp = int_temp // // P_hi = s_Y * P_hi // A_hi = s_Y * A_hi // (p0) fma.s1 Res_hi = sigma, A_hi, P_hi nop.i 999 ;; } { .mfi nop.m 999 (p0) fclass.m.unc p6,p0 = A_lo, 0x007 nop.i 999 ;; } { .mfi nop.m 999 (p6) mov A_lo = tmp nop.i 999 } { .mfi nop.m 999 // // Res_hi = P_hi + sigma * A_hi // (p0) fsub.s1 tmp = P_hi, Res_hi nop.i 999 ;; } { .mfi nop.m 999 // // tmp = P_hi - Res_hi // (p0) fma.s1 tmp = A_hi, sigma, tmp nop.i 999 } { .mfi nop.m 999 (p0) fma.s1 sigma = A_lo, sigma, P_lo nop.i 999 ;; } { .mfi nop.m 999 // // tmp = sigma * A_hi + tmp // sigma = A_lo * sigma + P_lo // (p0) fma.s1 Res_lo = s_Y, sigma, tmp nop.i 999 ;; } { .mfb nop.m 999 // // Res_lo = s_Y * sigma + tmp // (p0) fadd.s0 Result = Res_lo, Res_hi br.ret.sptk b0 ;; } L(ATANL_NATVAL): L(ATANL_UNSUPPORTED): L(ATANL_NAN): { .mfb nop.m 999 (p0) fmpy.s0 Result = ArgX,ArgY (p0) br.ret.sptk b0 ;; } L(ATANL_SPECIAL_HANDLING): { .mfi nop.m 999 (p0) fcmp.eq.s0 p0, p6 = f1, ArgY_orig nop.i 999 } { .mfi nop.m 999 (p0) fcmp.eq.s0 p0, p5 = f1, ArgX_orig nop.i 999 ;; } { .mfi nop.m 999 (p0) fclass.m.unc p6, p7 = ArgY, 0x007 nop.i 999 } { .mlx nop.m 999 (p0) movl special = 992 } ;; { .mmi nop.m 999 (p0) addl table_ptr1 = @ltoff(Constants_atan#), gp nop.i 999 } ;; { .mmi ld8 table_ptr1 = [table_ptr1] nop.m 999 nop.i 999 } ;; { .mib (p0) add table_ptr1 = table_ptr1, special nop.i 999 (p7) br.cond.spnt L(ATANL_ArgY_Not_ZERO) ;; } { .mmf (p0) ldfd Result = [table_ptr1], 8 nop.m 999 (p6) fclass.m.unc p14, p0 = ArgX, 0x035 ;; } { .mmf nop.m 999 (p0) ldfd Result_lo = [table_ptr1], -8 (p6) fclass.m.unc p15, p0 = ArgX, 0x036 ;; } { .mfi nop.m 999 (p14) fmerge.s Result = ArgY, f0 nop.i 999 } { .mfi nop.m 999 (p6) fclass.m.unc p13, p0 = ArgX, 0x007 nop.i 999 ;; } { .mfi nop.m 999 (p14) fmerge.s Result_lo = ArgY, f0 nop.i 999 ;; } { .mfi (p13) mov GR_Parameter_TAG = 36 nop.f 999 nop.i 999 ;; } { .mfi nop.m 999 // // Return sign_Y * 0 when ArgX > +0 // (p15) fmerge.s Result = ArgY, Result nop.i 999 ;; } { .mfi nop.m 999 (p15) fmerge.s Result_lo = ArgY, Result_lo nop.i 999 ;; } { .mfb nop.m 999 // // Return sign_Y * 0 when ArgX < -0 // (p0) fadd.s0 Result = Result, Result_lo (p13) br.cond.spnt __libm_error_region ;; } { .mib nop.m 999 nop.i 999 // // Call error support funciton for atan(0,0) // (p0) br.ret.sptk b0 ;; } L(ATANL_ArgY_Not_ZERO): { .mfi nop.m 999 (p0) fclass.m.unc p9, p10 = ArgY, 0x023 nop.i 999 ;; } { .mib nop.m 999 nop.i 999 (p10) br.cond.spnt L(ATANL_ArgY_Not_INF) ;; } { .mfi nop.m 999 (p9) fclass.m.unc p6, p0 = ArgX, 0x017 nop.i 999 } { .mfi nop.m 999 (p9) fclass.m.unc p7, p0 = ArgX, 0x021 nop.i 999 ;; } { .mfi nop.m 999 (p9) fclass.m.unc p8, p0 = ArgX, 0x022 nop.i 999 ;; } { .mmi (p6) add table_ptr1 = 16, table_ptr1 ;; (p0) ldfd Result = [table_ptr1], 8 nop.i 999 ;; } { .mfi (p0) ldfd Result_lo = [table_ptr1], -8 nop.f 999 nop.i 999 ;; } { .mfi nop.m 999 (p6) fmerge.s Result = ArgY, Result nop.i 999 ;; } { .mfi nop.m 999 (p6) fmerge.s Result_lo = ArgY, Result_lo nop.i 999 ;; } { .mfb nop.m 999 (p6) fadd.s0 Result = Result, Result_lo (p6) br.ret.sptk b0 ;; } // // Load PI/2 and adjust its sign. // Return +PI/2 when ArgY = +Inf and ArgX = +/-0 or normal // Return -PI/2 when ArgY = -Inf and ArgX = +/-0 or normal // { .mmi (p7) add table_ptr1 = 32, table_ptr1 ;; (p7) ldfd Result = [table_ptr1], 8 nop.i 999 ;; } { .mfi (p7) ldfd Result_lo = [table_ptr1], -8 nop.f 999 nop.i 999 ;; } { .mfi nop.m 999 (p7) fmerge.s Result = ArgY, Result nop.i 999 ;; } { .mfi nop.m 999 (p7) fmerge.s Result_lo = ArgY, Result_lo nop.i 999 ;; } { .mfb nop.m 999 (p7) fadd.s0 Result = Result, Result_lo (p7) br.ret.sptk b0 ;; } // // Load PI/4 and adjust its sign. // Return +PI/4 when ArgY = +Inf and ArgX = +Inf // Return -PI/4 when ArgY = -Inf and ArgX = +Inf // { .mmi (p8) add table_ptr1 = 48, table_ptr1 ;; (p8) ldfd Result = [table_ptr1], 8 nop.i 999 ;; } { .mfi (p8) ldfd Result_lo = [table_ptr1], -8 nop.f 999 nop.i 999 ;; } { .mfi nop.m 999 (p8) fmerge.s Result = ArgY, Result nop.i 999 ;; } { .mfi nop.m 999 (p8) fmerge.s Result_lo = ArgY, Result_lo nop.i 999 ;; } { .mfb nop.m 999 (p8) fadd.s0 Result = Result, Result_lo (p8) br.ret.sptk b0 ;; } L(ATANL_ArgY_Not_INF): { .mfi nop.m 999 // // Load PI/4 and adjust its sign. // Return +3PI/4 when ArgY = +Inf and ArgX = -Inf // Return -3PI/4 when ArgY = -Inf and ArgX = -Inf // (p0) fclass.m.unc p6, p0 = ArgX, 0x007 nop.i 999 } { .mfi nop.m 999 (p0) fclass.m.unc p7, p0 = ArgX, 0x021 nop.i 999 ;; } { .mfi nop.m 999 (p0) fclass.m.unc p8, p0 = ArgX, 0x022 nop.i 999 ;; } { .mmi (p6) add table_ptr1 = 16, table_ptr1 ;; (p6) ldfd Result = [table_ptr1], 8 nop.i 999 ;; } { .mfi (p6) ldfd Result_lo = [table_ptr1], -8 nop.f 999 nop.i 999 ;; } { .mfi nop.m 999 (p6) fmerge.s Result = ArgY, Result nop.i 999 ;; } { .mfi nop.m 999 (p6) fmerge.s Result_lo = ArgY, Result_lo nop.i 999 ;; } { .mfb nop.m 999 (p6) fadd.s0 Result = Result, Result_lo (p6) br.ret.spnt b0 ;; } { .mfi nop.m 999 // // return = sign_Y * PI/2 when ArgX = 0 // (p7) fmerge.s Result = ArgY, f0 nop.i 999 ;; } { .mfb nop.m 999 (p7) fnorm.s0 Result = Result (p7) br.ret.spnt b0 ;; } // // return = sign_Y * 0 when ArgX = Inf // { .mmi (p8) ldfd Result = [table_ptr1], 8 ;; (p8) ldfd Result_lo = [table_ptr1], -8 nop.i 999 ;; } { .mfi nop.m 999 (p8) fmerge.s Result = ArgY, Result nop.i 999 ;; } { .mfi nop.m 999 (p8) fmerge.s Result_lo = ArgY, Result_lo nop.i 999 ;; } { .mfb nop.m 999 (p8) fadd.s0 Result = Result, Result_lo (p8) br.ret.sptk b0 ;; } // // return = sign_Y * PI when ArgX = -Inf // .endp atan2l ASM_SIZE_DIRECTIVE(atan2l) ASM_SIZE_DIRECTIVE(__atan2l) ASM_SIZE_DIRECTIVE(__ieee754_atan2l) .proc __libm_error_region __libm_error_region: .prologue { .mfi add GR_Parameter_Y=-32,sp // Parameter 2 value nop.f 0 .save ar.pfs,GR_SAVE_PFS mov GR_SAVE_PFS=ar.pfs // Save ar.pfs } { .mfi .fframe 64 add sp=-64,sp // Create new stack nop.f 0 mov GR_SAVE_GP=gp // Save gp };; { .mmi stfe [GR_Parameter_Y] = FR_Y,16 // Save Parameter 2 on stack add GR_Parameter_X = 16,sp // Parameter 1 address .save b0, GR_SAVE_B0 mov GR_SAVE_B0=b0 // Save b0 };; .body { .mib stfe [GR_Parameter_X] = FR_X // Store Parameter 1 on stack add GR_Parameter_RESULT = 0,GR_Parameter_Y nop.b 0 // Parameter 3 address } { .mib stfe [GR_Parameter_Y] = FR_RESULT // Store Parameter 3 on stack add GR_Parameter_Y = -16,GR_Parameter_Y br.call.sptk b0=__libm_error_support# // Call error handling function };; { .mmi nop.m 0 nop.m 0 add GR_Parameter_RESULT = 48,sp };; { .mmi ldfe f8 = [GR_Parameter_RESULT] // Get return result off stack .restore sp add sp = 64,sp // Restore stack pointer mov b0 = GR_SAVE_B0 // Restore return address };; { .mib mov gp = GR_SAVE_GP // Restore gp mov ar.pfs = GR_SAVE_PFS // Restore ar.pfs br.ret.sptk b0 // Return };; .endp __libm_error_region ASM_SIZE_DIRECTIVE(__libm_error_region) .type __libm_error_support#,@function .global __libm_error_support#