https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126053
--- Comment #3 from Reshma Roy <Reshma.Roy at amd dot com> ---
The revised reproducer lies in the same path as the benchmark now,
but the test case ends up to be a bit longer. Added the preprocessed input as
an attachment.
Revised reduced test case
---------------------------
typedef double BigReal;
struct Position { BigReal x, y, z; };
struct CompAtom {
Position position;
float charge;
short vdwType;
unsigned char partition;
unsigned char nonbondedGroupSize;
};
struct Force { BigReal x, y, z; };
struct SimParams { BigReal offset_x, offset_y, offset_z; };
enum { vXX, vXY, vXZ, vYY, vYZ, vZZ,
fvXX, fvXY, fvXZ, fvYY, fvYZ, fvZZ };
# 65 "namd557_pair_fail.cpp"
void calc_pair_energy_fullelect(
const CompAtom * __restrict p_0,
const CompAtom * __restrict p_1,
const SimParams * __restrict params,
const int * __restrict pairlist_n,
const int * __restrict pairlist_m,
const int * __restrict npair_n_list,
const int * __restrict npair_m_list,
int i_upper,
const BigReal * __restrict force_r_vals,
Force * __restrict f_0,
Force * __restrict f_1,
BigReal * __restrict reduction)
{
BigReal virial_xx = 0, virial_xy = 0, virial_xz = 0;
BigReal virial_yy = 0, virial_yz = 0, virial_zz = 0;
BigReal fullElectVirial_xx = 0, fullElectVirial_xy = 0, fullElectVirial_xz
= 0;
BigReal fullElectVirial_yy = 0, fullElectVirial_yz = 0, fullElectVirial_zz
= 0;
int pn = 0, pm = 0;
for (int i = 0; i < i_upper; ++i) {
const CompAtom *p_i = p_0 + i;
const BigReal p_i_x = params->offset_x + p_i->position.x;
const BigReal p_i_y = params->offset_y + p_i->position.y;
const BigReal p_i_z = params->offset_z + p_i->position.z;
BigReal f_i_x = 0, f_i_y = 0, f_i_z = 0;
{
const int npairi = npair_n_list[i];
const int *pli = pairlist_n + pn;
const BigReal *fr = force_r_vals + pn;
for (int k = 0; k < npairi; ++k) { const int j = pli[k]; const
CompAtom *p_j = p_1 + j; Force *f_j = f_1 + j; const BigReal p_ij_x = p_i_x -
p_j->position.x; const BigReal p_ij_y = p_i_y - p_j->position.y; const BigReal
p_ij_z = p_i_z - p_j->position.z; const BigReal force_r = fr[k]; BigReal tmp_x
= force_r * p_ij_x; virial_xx += tmp_x * p_ij_x; virial_xy += tmp_x * p_ij_y;
virial_xz += tmp_x * p_ij_z; f_i_x += tmp_x; f_j->x -= tmp_x; BigReal tmp_y =
force_r * p_ij_y; virial_yy += tmp_y * p_ij_y; virial_yz += tmp_y * p_ij_z;
f_i_y += tmp_y; f_j->y -= tmp_y; BigReal tmp_z = force_r * p_ij_z; virial_zz +=
tmp_z * p_ij_z; f_i_z += tmp_z; f_j->z -= tmp_z; }
pn += npairi;
}
{
const int npairi = npair_m_list[i];
const int *pli = pairlist_m + pm;
const BigReal *fr = force_r_vals + pm;
for (int k = 0; k < npairi; ++k) { const int j = pli[k]; const
CompAtom *p_j = p_1 + j; Force *f_j = f_1 + j; const BigReal p_ij_x = p_i_x -
p_j->position.x; const BigReal p_ij_y = p_i_y - p_j->position.y; const BigReal
p_ij_z = p_i_z - p_j->position.z; const BigReal force_r = fr[k]; BigReal tmp_x
= force_r * p_ij_x; virial_xx += tmp_x * p_ij_x; virial_xy += tmp_x * p_ij_y;
virial_xz += tmp_x * p_ij_z; f_i_x += tmp_x; f_j->x -= tmp_x; BigReal tmp_y =
force_r * p_ij_y; virial_yy += tmp_y * p_ij_y; virial_yz += tmp_y * p_ij_z;
f_i_y += tmp_y; f_j->y -= tmp_y; BigReal tmp_z = force_r * p_ij_z; virial_zz +=
tmp_z * p_ij_z; f_i_z += tmp_z; f_j->z -= tmp_z; }
pm += npairi;
}
f_0[i].x += f_i_x;
f_0[i].y += f_i_y;
f_0[i].z += f_i_z;
}
reduction[vXX] += virial_xx;
reduction[vXY] += virial_xy;
reduction[vXZ] += virial_xz;
reduction[vYY] += virial_yy;
reduction[vYZ] += virial_yz;
reduction[vZZ] += virial_zz;
reduction[fvXX] += fullElectVirial_xx;
reduction[fvXY] += fullElectVirial_xy;
reduction[fvXZ] += fullElectVirial_xz;
reduction[fvYY] += fullElectVirial_yy;
reduction[fvYZ] += fullElectVirial_yz;
reduction[fvZZ] += fullElectVirial_zz;
}
GCC reports two messages on the scatter sites in both the reduced testcase and
NAMD (Base2.h:361/:540).
SLP dump on the scatter sites (macro expansion at lines 97 and 104):
namd557_pair_fail.cpp:97:442: missed: failed: evolution of base is not affine.
namd557_pair_fail.cpp:97:442: missed: Cannot determine insertion place for lane
extract
namd557_pair_fail.cpp:104:442: ... <same pair>
It looks like SLP packs lane 0 = x and lane 1 = y of the same pair into one
vector, but looks like there are other statements that consume those values as
scalars and are not in the scatter SLP tree: like virial_xx += tmp_x * p_ij_x;
...
(similar to benchmark source ComputeNonbondedBase2.h:354-372) which might be
the reason for the failure.
