From 5590164fe2db998f100b69d6c26b7b2d404fb273 Mon Sep 17 00:00:00 2001 From: Angelika Schwarz Date: Sun, 11 Jun 2023 18:34:42 +0200 Subject: [PATCH] Remove outdated warning when guard digits are missing This comment is irrelevant on any machine that realizes IEEE 754. --- SRC/cgelsd.f | 6 ------ SRC/cgesdd.f | 6 ------ SRC/chbevd.f | 6 ------ SRC/chbevd_2stage.f | 6 ------ SRC/chbgvd.f | 6 ------ SRC/cheevd.f | 6 ------ SRC/cheevd_2stage.f | 6 ------ SRC/chegvd.f | 6 ------ SRC/chpevd.f | 6 ------ SRC/chpgvd.f | 6 ------ SRC/clalsd.f | 6 ------ SRC/cstedc.f | 6 ------ SRC/dbdsdc.f | 7 ------- SRC/dgelsd.f | 6 ------ SRC/dgesdd.f | 6 ------ SRC/dlalsd.f | 6 ------ SRC/dsbevd.f | 6 ------ SRC/dsbevd_2stage.f | 6 ------ SRC/dsbgvd.f | 6 ------ SRC/dspevd.f | 6 ------ SRC/dspgvd.f | 6 ------ SRC/dstedc.f | 6 ------ SRC/dstevd.f | 6 ------ SRC/dsyevd.f | 7 ------- SRC/dsyevd_2stage.f | 6 ------ SRC/dsygvd.f | 6 ------ SRC/sbdsdc.f | 7 ------- SRC/sgelsd.f | 6 ------ SRC/sgesdd.f | 6 ------ SRC/slalsd.f | 6 ------ SRC/ssbevd.f | 6 ------ SRC/ssbevd_2stage.f | 6 ------ SRC/ssbgvd.f | 6 ------ SRC/sspevd.f | 6 ------ SRC/sspgvd.f | 6 ------ SRC/sstedc.f | 6 ------ SRC/sstevd.f | 6 ------ SRC/ssyevd.f | 7 ------- SRC/ssyevd_2stage.f | 6 ------ SRC/ssygvd.f | 6 ------ SRC/zgelsd.f | 6 ------ SRC/zgesdd.f | 6 ------ SRC/zhbevd.f | 6 ------ SRC/zhbevd_2stage.f | 6 ------ SRC/zhbgvd.f | 6 ------ SRC/zheevd.f | 6 ------ SRC/zheevd_2stage.f | 6 ------ SRC/zhegvd.f | 6 ------ SRC/zhpevd.f | 6 ------ SRC/zhpgvd.f | 6 ------ SRC/zlalsd.f | 6 ------ SRC/zstedc.f | 6 ------ 52 files changed, 316 deletions(-) diff --git a/SRC/cgelsd.f b/SRC/cgelsd.f index 93b81aa438..4dd50173a5 100644 --- a/SRC/cgelsd.f +++ b/SRC/cgelsd.f @@ -60,12 +60,6 @@ *> singular values which are less than RCOND times the largest singular *> value. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/cgesdd.f b/SRC/cgesdd.f index 12194dc7f6..9adc254e00 100644 --- a/SRC/cgesdd.f +++ b/SRC/cgesdd.f @@ -53,12 +53,6 @@ *> *> Note that the routine returns VT = V**H, not V. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/chbevd.f b/SRC/chbevd.f index 1598f4de5e..de33c9039c 100644 --- a/SRC/chbevd.f +++ b/SRC/chbevd.f @@ -41,12 +41,6 @@ *> a complex Hermitian band matrix A. If eigenvectors are desired, it *> uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/chbevd_2stage.f b/SRC/chbevd_2stage.f index 340c546e8c..3c9c8ecc02 100644 --- a/SRC/chbevd_2stage.f +++ b/SRC/chbevd_2stage.f @@ -47,12 +47,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it *> uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/chbgvd.f b/SRC/chbgvd.f index c4ad20753f..6550063708 100644 --- a/SRC/chbgvd.f +++ b/SRC/chbgvd.f @@ -46,12 +46,6 @@ *> and banded, and B is also positive definite. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/cheevd.f b/SRC/cheevd.f index 2ddf74b985..dce0b20834 100644 --- a/SRC/cheevd.f +++ b/SRC/cheevd.f @@ -41,12 +41,6 @@ *> complex Hermitian matrix A. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/cheevd_2stage.f b/SRC/cheevd_2stage.f index 830e13d301..a0e8843aee 100644 --- a/SRC/cheevd_2stage.f +++ b/SRC/cheevd_2stage.f @@ -46,12 +46,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/chegvd.f b/SRC/chegvd.f index c96f011af8..4edc36f2ad 100644 --- a/SRC/chegvd.f +++ b/SRC/chegvd.f @@ -43,12 +43,6 @@ *> B are assumed to be Hermitian and B is also positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/chpevd.f b/SRC/chpevd.f index c44462394e..06d01064df 100644 --- a/SRC/chpevd.f +++ b/SRC/chpevd.f @@ -41,12 +41,6 @@ *> a complex Hermitian matrix A in packed storage. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/chpgvd.f b/SRC/chpgvd.f index 5c9e417d3b..c24ca13609 100644 --- a/SRC/chpgvd.f +++ b/SRC/chpgvd.f @@ -44,12 +44,6 @@ *> positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/clalsd.f b/SRC/clalsd.f index a2da9a9257..bdd6b31c58 100644 --- a/SRC/clalsd.f +++ b/SRC/clalsd.f @@ -48,12 +48,6 @@ *> problem; in this case a minimum norm solution is returned. *> The actual singular values are returned in D in ascending order. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray XMP, Cray YMP, Cray C 90, or Cray 2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/cstedc.f b/SRC/cstedc.f index a57d9eaef1..77a4ec3be4 100644 --- a/SRC/cstedc.f +++ b/SRC/cstedc.f @@ -43,12 +43,6 @@ *> be found if CHETRD or CHPTRD or CHBTRD has been used to reduce this *> matrix to tridiagonal form. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or Cray-2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. See SLAED3 for details. *> \endverbatim * * Arguments: diff --git a/SRC/dbdsdc.f b/SRC/dbdsdc.f index 99fe82296d..4b6c3e6943 100644 --- a/SRC/dbdsdc.f +++ b/SRC/dbdsdc.f @@ -45,13 +45,6 @@ *> respectively. DBDSDC can be used to compute all singular values, *> and optionally, singular vectors or singular vectors in compact form. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or Cray-2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. See DLASD3 for details. -*> *> The code currently calls DLASDQ if singular values only are desired. *> However, it can be slightly modified to compute singular values *> using the divide and conquer method. diff --git a/SRC/dgelsd.f b/SRC/dgelsd.f index 46de3d7fbd..9e46abc92a 100644 --- a/SRC/dgelsd.f +++ b/SRC/dgelsd.f @@ -59,12 +59,6 @@ *> singular values which are less than RCOND times the largest singular *> value. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dgesdd.f b/SRC/dgesdd.f index 27a6c1f8c0..d49524bd2c 100644 --- a/SRC/dgesdd.f +++ b/SRC/dgesdd.f @@ -55,12 +55,6 @@ *> *> Note that the routine returns VT = V**T, not V. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dlalsd.f b/SRC/dlalsd.f index d22c45dc6e..706ac4c90e 100644 --- a/SRC/dlalsd.f +++ b/SRC/dlalsd.f @@ -47,12 +47,6 @@ *> problem; in this case a minimum norm solution is returned. *> The actual singular values are returned in D in ascending order. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray XMP, Cray YMP, Cray C 90, or Cray 2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dsbevd.f b/SRC/dsbevd.f index 3eb4ed8df1..350c0a9f08 100644 --- a/SRC/dsbevd.f +++ b/SRC/dsbevd.f @@ -40,12 +40,6 @@ *> a real symmetric band matrix A. If eigenvectors are desired, it uses *> a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dsbevd_2stage.f b/SRC/dsbevd_2stage.f index 45a64b4781..82997c8502 100644 --- a/SRC/dsbevd_2stage.f +++ b/SRC/dsbevd_2stage.f @@ -45,12 +45,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it uses *> a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dsbgvd.f b/SRC/dsbgvd.f index 30b0166110..0ab3177ace 100644 --- a/SRC/dsbgvd.f +++ b/SRC/dsbgvd.f @@ -43,12 +43,6 @@ *> banded, and B is also positive definite. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dspevd.f b/SRC/dspevd.f index d9d6c89177..05aa91b03a 100644 --- a/SRC/dspevd.f +++ b/SRC/dspevd.f @@ -40,12 +40,6 @@ *> of a real symmetric matrix A in packed storage. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dspgvd.f b/SRC/dspgvd.f index ec3cdc1ac6..24c2309c34 100644 --- a/SRC/dspgvd.f +++ b/SRC/dspgvd.f @@ -44,12 +44,6 @@ *> positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dstedc.f b/SRC/dstedc.f index 2ed84afaac..6d533664bd 100644 --- a/SRC/dstedc.f +++ b/SRC/dstedc.f @@ -42,12 +42,6 @@ *> found if DSYTRD or DSPTRD or DSBTRD has been used to reduce this *> matrix to tridiagonal form. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or Cray-2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. See DLAED3 for details. *> \endverbatim * * Arguments: diff --git a/SRC/dstevd.f b/SRC/dstevd.f index 507f39b2b6..54717df3d4 100644 --- a/SRC/dstevd.f +++ b/SRC/dstevd.f @@ -40,12 +40,6 @@ *> real symmetric tridiagonal matrix. If eigenvectors are desired, it *> uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dsyevd.f b/SRC/dsyevd.f index eaaecd8d98..b27f4cdc7a 100644 --- a/SRC/dsyevd.f +++ b/SRC/dsyevd.f @@ -40,13 +40,6 @@ *> real symmetric matrix A. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. -*> *> Because of large use of BLAS of level 3, DSYEVD needs N**2 more *> workspace than DSYEVX. *> \endverbatim diff --git a/SRC/dsyevd_2stage.f b/SRC/dsyevd_2stage.f index 0eae8ad062..d5a68c35df 100644 --- a/SRC/dsyevd_2stage.f +++ b/SRC/dsyevd_2stage.f @@ -45,12 +45,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/dsygvd.f b/SRC/dsygvd.f index d6682d4e5c..41a384c806 100644 --- a/SRC/dsygvd.f +++ b/SRC/dsygvd.f @@ -42,12 +42,6 @@ *> B are assumed to be symmetric and B is also positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/sbdsdc.f b/SRC/sbdsdc.f index 18a4044979..2a6cc99708 100644 --- a/SRC/sbdsdc.f +++ b/SRC/sbdsdc.f @@ -45,13 +45,6 @@ *> respectively. SBDSDC can be used to compute all singular values, *> and optionally, singular vectors or singular vectors in compact form. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or Cray-2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. See SLASD3 for details. -*> *> The code currently calls SLASDQ if singular values only are desired. *> However, it can be slightly modified to compute singular values *> using the divide and conquer method. diff --git a/SRC/sgelsd.f b/SRC/sgelsd.f index a680472e1a..0abd158498 100644 --- a/SRC/sgelsd.f +++ b/SRC/sgelsd.f @@ -59,12 +59,6 @@ *> singular values which are less than RCOND times the largest singular *> value. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/sgesdd.f b/SRC/sgesdd.f index 4f9100d43c..71af7b3720 100644 --- a/SRC/sgesdd.f +++ b/SRC/sgesdd.f @@ -55,12 +55,6 @@ *> *> Note that the routine returns VT = V**T, not V. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/slalsd.f b/SRC/slalsd.f index 2197f728e0..9943a52d9e 100644 --- a/SRC/slalsd.f +++ b/SRC/slalsd.f @@ -47,12 +47,6 @@ *> problem; in this case a minimum norm solution is returned. *> The actual singular values are returned in D in ascending order. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray XMP, Cray YMP, Cray C 90, or Cray 2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/ssbevd.f b/SRC/ssbevd.f index bcf14ce85e..e87f9a0304 100644 --- a/SRC/ssbevd.f +++ b/SRC/ssbevd.f @@ -40,12 +40,6 @@ *> a real symmetric band matrix A. If eigenvectors are desired, it uses *> a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/ssbevd_2stage.f b/SRC/ssbevd_2stage.f index 9687ee0243..014bade48c 100644 --- a/SRC/ssbevd_2stage.f +++ b/SRC/ssbevd_2stage.f @@ -45,12 +45,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it uses *> a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/ssbgvd.f b/SRC/ssbgvd.f index 6dd1fe952b..7c21ee455c 100644 --- a/SRC/ssbgvd.f +++ b/SRC/ssbgvd.f @@ -43,12 +43,6 @@ *> banded, and B is also positive definite. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/sspevd.f b/SRC/sspevd.f index 56329da341..0872e95acd 100644 --- a/SRC/sspevd.f +++ b/SRC/sspevd.f @@ -40,12 +40,6 @@ *> of a real symmetric matrix A in packed storage. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/sspgvd.f b/SRC/sspgvd.f index 8ce2311fa0..1a88365f2a 100644 --- a/SRC/sspgvd.f +++ b/SRC/sspgvd.f @@ -44,12 +44,6 @@ *> positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/sstedc.f b/SRC/sstedc.f index 925b03422e..61e3c2fda7 100644 --- a/SRC/sstedc.f +++ b/SRC/sstedc.f @@ -42,12 +42,6 @@ *> found if SSYTRD or SSPTRD or SSBTRD has been used to reduce this *> matrix to tridiagonal form. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or Cray-2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. See SLAED3 for details. *> \endverbatim * * Arguments: diff --git a/SRC/sstevd.f b/SRC/sstevd.f index bc5b5aaab7..218af8c768 100644 --- a/SRC/sstevd.f +++ b/SRC/sstevd.f @@ -40,12 +40,6 @@ *> real symmetric tridiagonal matrix. If eigenvectors are desired, it *> uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/ssyevd.f b/SRC/ssyevd.f index ac0d0284d3..ee0e33384e 100644 --- a/SRC/ssyevd.f +++ b/SRC/ssyevd.f @@ -40,13 +40,6 @@ *> real symmetric matrix A. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. -*> *> Because of large use of BLAS of level 3, SSYEVD needs N**2 more *> workspace than SSYEVX. *> \endverbatim diff --git a/SRC/ssyevd_2stage.f b/SRC/ssyevd_2stage.f index f3fde6b4a1..e63e280a79 100644 --- a/SRC/ssyevd_2stage.f +++ b/SRC/ssyevd_2stage.f @@ -45,12 +45,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/ssygvd.f b/SRC/ssygvd.f index 79f12a6f9a..3c8bd2a0ec 100644 --- a/SRC/ssygvd.f +++ b/SRC/ssygvd.f @@ -42,12 +42,6 @@ *> B are assumed to be symmetric and B is also positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zgelsd.f b/SRC/zgelsd.f index 15ca42300f..dad746564b 100644 --- a/SRC/zgelsd.f +++ b/SRC/zgelsd.f @@ -60,12 +60,6 @@ *> singular values which are less than RCOND times the largest singular *> value. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zgesdd.f b/SRC/zgesdd.f index a3e6408bf8..6064510b0a 100644 --- a/SRC/zgesdd.f +++ b/SRC/zgesdd.f @@ -53,12 +53,6 @@ *> *> Note that the routine returns VT = V**H, not V. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zhbevd.f b/SRC/zhbevd.f index 0db5515409..be9f015560 100644 --- a/SRC/zhbevd.f +++ b/SRC/zhbevd.f @@ -41,12 +41,6 @@ *> a complex Hermitian band matrix A. If eigenvectors are desired, it *> uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zhbevd_2stage.f b/SRC/zhbevd_2stage.f index 4522d5e796..e32c7125ca 100644 --- a/SRC/zhbevd_2stage.f +++ b/SRC/zhbevd_2stage.f @@ -47,12 +47,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it *> uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zhbgvd.f b/SRC/zhbgvd.f index b0664750e7..4bd02168d4 100644 --- a/SRC/zhbgvd.f +++ b/SRC/zhbgvd.f @@ -46,12 +46,6 @@ *> and banded, and B is also positive definite. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zheevd.f b/SRC/zheevd.f index 7f58c7f726..ba52f9e723 100644 --- a/SRC/zheevd.f +++ b/SRC/zheevd.f @@ -41,12 +41,6 @@ *> complex Hermitian matrix A. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zheevd_2stage.f b/SRC/zheevd_2stage.f index 9859b0d67f..e697a98237 100644 --- a/SRC/zheevd_2stage.f +++ b/SRC/zheevd_2stage.f @@ -46,12 +46,6 @@ *> the reduction to tridiagonal. If eigenvectors are desired, it uses a *> divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zhegvd.f b/SRC/zhegvd.f index 2c3586517c..c9ff55e3d2 100644 --- a/SRC/zhegvd.f +++ b/SRC/zhegvd.f @@ -43,12 +43,6 @@ *> B are assumed to be Hermitian and B is also positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zhpevd.f b/SRC/zhpevd.f index 7625c8fe81..5260aaf14a 100644 --- a/SRC/zhpevd.f +++ b/SRC/zhpevd.f @@ -41,12 +41,6 @@ *> a complex Hermitian matrix A in packed storage. If eigenvectors are *> desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zhpgvd.f b/SRC/zhpgvd.f index e9688f0c7a..dfe92067cf 100644 --- a/SRC/zhpgvd.f +++ b/SRC/zhpgvd.f @@ -44,12 +44,6 @@ *> positive definite. *> If eigenvectors are desired, it uses a divide and conquer algorithm. *> -*> The divide and conquer algorithm makes very mild assumptions about -*> floating point arithmetic. It will work on machines with a guard -*> digit in add/subtract, or on those binary machines without guard -*> digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or -*> Cray-2. It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zlalsd.f b/SRC/zlalsd.f index dca308e561..1d7358aa97 100644 --- a/SRC/zlalsd.f +++ b/SRC/zlalsd.f @@ -48,12 +48,6 @@ *> problem; in this case a minimum norm solution is returned. *> The actual singular values are returned in D in ascending order. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray XMP, Cray YMP, Cray C 90, or Cray 2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. *> \endverbatim * * Arguments: diff --git a/SRC/zstedc.f b/SRC/zstedc.f index 74d390af7e..e62063a19e 100644 --- a/SRC/zstedc.f +++ b/SRC/zstedc.f @@ -43,12 +43,6 @@ *> be found if ZHETRD or ZHPTRD or ZHBTRD has been used to reduce this *> matrix to tridiagonal form. *> -*> This code makes very mild assumptions about floating point -*> arithmetic. It will work on machines with a guard digit in -*> add/subtract, or on those binary machines without guard digits -*> which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or Cray-2. -*> It could conceivably fail on hexadecimal or decimal machines -*> without guard digits, but we know of none. See DLAED3 for details. *> \endverbatim * * Arguments: