diff --git a/lib/gpu/lal_born_coul_long_cs.cu b/lib/gpu/lal_born_coul_long_cs.cu
index 69503ba420..6f04fcea94 100644
--- a/lib/gpu/lal_born_coul_long_cs.cu
+++ b/lib/gpu/lal_born_coul_long_cs.cu
@@ -109,7 +109,7 @@ __kernel void k_born_coul_long_cs(const __global numtyp4 *restrict x_,
       numtyp rsq = delx*delx+dely*dely+delz*delz;
 
       int mtype=itype*lj_types+jtype;
-      if (rsq<cutsq_sigma[mtype].x) { // cutsq 
+      if (rsq<cutsq_sigma[mtype].x) { // cutsq
         numtyp forcecoul,forceborn,force,r6inv,prefactor,_erfc,rexp;
 
         rsq += EPSILON; // Add Epsilon for case: r = 0; Interaction must be removed by special bond;
@@ -250,7 +250,7 @@ __kernel void k_born_coul_long_cs_fast(const __global numtyp4 *restrict x_,
       numtyp delz = ix.z-jx.z;
       numtyp rsq = delx*delx+dely*dely+delz*delz;
 
-      if (rsq<cutsq_sigma[mtype].x) { // cutsq 
+      if (rsq<cutsq_sigma[mtype].x) { // cutsq
         numtyp forcecoul,forceborn,force,r6inv,prefactor,_erfc,rexp;
 
         rsq += EPSILON; // Add Epsilon for case: r = 0; Interaction must be removed by special bond;
diff --git a/lib/gpu/lal_ufm.cu b/lib/gpu/lal_ufm.cu
index 33d0f3c956..03d1e85bdf 100644
--- a/lib/gpu/lal_ufm.cu
+++ b/lib/gpu/lal_ufm.cu
@@ -10,7 +10,7 @@
     This file is part of the LAMMPS Accelerator Library (LAMMPS_AL)
  __________________________________________________________________________
 
-    begin                : 
+    begin                :
     email                : pl.rodolfo@gmail.com
                            dekoning@ifi.unicamp.br
  ***************************************************************************/
@@ -31,10 +31,10 @@ __kernel void k_ufm(const __global numtyp4 *restrict x_,
                    const __global numtyp4 *restrict uf3,
                    const int lj_types,
                    const __global numtyp *restrict sp_lj,
-                   const __global int * dev_nbor, 
-                   const __global int * dev_packed, 
-                   __global acctyp4 *restrict ans, 
-                   __global acctyp *restrict engv, 
+                   const __global int * dev_nbor,
+                   const __global int * dev_packed,
+                   __global acctyp4 *restrict ans,
+                   __global acctyp *restrict engv,
                    const int eflag, const int vflag, const int inum,
                    const int nbor_pitch, const int t_per_atom) {
   int tid, ii, offset;
@@ -46,19 +46,19 @@ __kernel void k_ufm(const __global numtyp4 *restrict x_,
   acctyp virial[6];
   for (int i=0; i<6; i++)
     virial[i]=(acctyp)0;
-  
+
   if (ii<inum) {
     int i, numj, nbor, nbor_end;
     __local int n_stride;
     nbor_info(dev_nbor,dev_packed,nbor_pitch,t_per_atom,ii,offset,i,numj,
               n_stride,nbor_end,nbor);
-  
+
     numtyp4 ix; fetch4(ix,i,pos_tex); //x_[i];
     int itype=ix.w;
 
     numtyp factor_lj;
     for ( ; nbor<nbor_end; nbor+=n_stride) {
-  
+
       int j=dev_packed[nbor];
       factor_lj = sp_lj[sbmask(j)];
       j &= NEIGHMASK;
@@ -71,12 +71,12 @@ __kernel void k_ufm(const __global numtyp4 *restrict x_,
       numtyp dely = ix.y-jx.y;
       numtyp delz = ix.z-jx.z;
       numtyp rsq = delx*delx+dely*dely+delz*delz;
- 
+
       int mtype=itype*lj_types+jtype;
       if (rsq<uf1[mtype].z) {
         numtyp expuf = exp(- rsq * uf1[mtype].y);
         numtyp force = factor_lj * uf1[mtype].x * expuf / (1.0 - expuf);
-      
+
         f.x += delx*force;
         f.y += dely*force;
         f.z += delz*force;
@@ -102,17 +102,17 @@ __kernel void k_ufm(const __global numtyp4 *restrict x_,
 
 __kernel void k_ufm_fast(const __global numtyp4 *restrict x_,
                         const __global numtyp4 *restrict uf1_in,
-                        const __global numtyp4 *restrict uf3_in, 
+                        const __global numtyp4 *restrict uf3_in,
                         const __global numtyp *restrict sp_lj_in,
-                        const __global int * dev_nbor, 
-                        const __global int * dev_packed, 
-                        __global acctyp4 *restrict ans, 
-                        __global acctyp *restrict engv, 
-                        const int eflag, const int vflag, const int inum, 
+                        const __global int * dev_nbor,
+                        const __global int * dev_packed,
+                        __global acctyp4 *restrict ans,
+                        __global acctyp *restrict engv,
+                        const int eflag, const int vflag, const int inum,
                         const int nbor_pitch, const int t_per_atom) {
   int tid, ii, offset;
   atom_info(t_per_atom,ii,tid,offset);
-  
+
   __local numtyp4 uf1[MAX_SHARED_TYPES*MAX_SHARED_TYPES];
   __local numtyp4 uf3[MAX_SHARED_TYPES*MAX_SHARED_TYPES];
   __local numtyp sp_lj[4];
@@ -123,7 +123,7 @@ __kernel void k_ufm_fast(const __global numtyp4 *restrict x_,
     if (eflag>0)
       uf3[tid]=uf3_in[tid];
   }
-  
+
   acctyp energy=(acctyp)0;
   acctyp4 f;
   f.x=(acctyp)0; f.y=(acctyp)0; f.z=(acctyp)0;
@@ -132,7 +132,7 @@ __kernel void k_ufm_fast(const __global numtyp4 *restrict x_,
     virial[i]=(acctyp)0;
 
   __syncthreads();
-  
+
   if (ii<inum) {
     int i, numj, nbor, nbor_end;
     __local int n_stride;
@@ -145,7 +145,7 @@ __kernel void k_ufm_fast(const __global numtyp4 *restrict x_,
 
     numtyp factor_lj;
     for ( ; nbor<nbor_end; nbor+=n_stride) {
-  
+
       int j=dev_packed[nbor];
       factor_lj = sp_lj[sbmask(j)];
       j &= NEIGHMASK;
@@ -158,11 +158,11 @@ __kernel void k_ufm_fast(const __global numtyp4 *restrict x_,
       numtyp dely = ix.y-jx.y;
       numtyp delz = ix.z-jx.z;
       numtyp rsq = delx*delx+dely*dely+delz*delz;
-        
+
       if (rsq<uf1[mtype].z) {
         numtyp expuf = exp(- rsq * uf1[mtype].y);
         numtyp force = factor_lj * uf1[mtype].x * expuf / (1.0 - expuf);
-      
+
         f.x += delx*force;
         f.y += dely*force;
         f.z += delz*force;
diff --git a/lib/gpu/lal_vashishta.cu b/lib/gpu/lal_vashishta.cu
index 07b7f212fc..25089c0c21 100644
--- a/lib/gpu/lal_vashishta.cu
+++ b/lib/gpu/lal_vashishta.cu
@@ -290,7 +290,7 @@ __kernel void k_vashishta(const __global numtyp4 *restrict x_,
 
         if (eflag>0)
           energy += (param3_bigh*reta+vc2-vc3-param3_bigw*r6inv-r*param3_dvrc+param3_c0);
-          
+
         if (vflag>0) {
           virial[0] += delx*delx*force;
           virial[1] += dely*dely*force;
@@ -471,13 +471,13 @@ __kernel void k_vashishta_three_center(const __global numtyp4 *restrict x_,
       numtyp rsq1 = delr1x*delr1x+delr1y*delr1y+delr1z*delr1z;
 
       int ijparam=elem2param[itype*nelements*nelements+jtype*nelements+jtype];
-      
+
       numtyp4 param4_ijparam; fetch4(param4_ijparam,ijparam,param4_tex);
       param_r0sq_ij=param4_ijparam.x;
       if (rsq1 > param_r0sq_ij) continue; // still keep this for neigh no and tpa > 1
       param_gamma_ij=param4_ijparam.y;
       param_r0_ij=param4_ijparam.w;
-      
+
       int nbor_k,k_end;
       if (dev_packed==dev_nbor) {
         nbor_k=nborj_start-offset_j+offset_k;
@@ -619,7 +619,7 @@ __kernel void k_vashishta_three_end(const __global numtyp4 *restrict x_,
 
       param_gamma_ij=param4_ijparam.y;
       param_r0_ij = param4_ijparam.w;
-      
+
       int nbor_k,numk;
       if (dev_nbor==dev_packed) {
         if (gpu_nbor) nbor_k=j+nbor_pitch;
@@ -665,14 +665,14 @@ __kernel void k_vashishta_three_end(const __global numtyp4 *restrict x_,
         if (rsq2 < param_r0sq_ik) {
           param_gamma_ik=param4_ikparam.y;
           param_r0_ik=param4_ikparam.w;
-          
+
           int ijkparam=elem2param[jtype*nelements*nelements+itype*nelements+ktype]; //jik
           numtyp4 param5_ijkparam; fetch4(param5_ijkparam,ijkparam,param5_tex);
           param_bigc_ijk=param5_ijkparam.x;
           param_costheta_ijk=param5_ijkparam.y;
           param_bigb_ijk=param5_ijkparam.z;
           param_big2b_ijk=param5_ijkparam.w;
-          
+
           numtyp fjx, fjy, fjz;
           //if (evatom==0) {
             threebody_half(delr1x,delr1y,delr1z);
@@ -774,7 +774,7 @@ __kernel void k_vashishta_three_end_vatom(const __global numtyp4 *restrict x_,
 
       param_gamma_ij=param4_ijparam.y;
       param_r0_ij=param4_ijparam.w;
-      
+
       int nbor_k,numk;
       if (dev_nbor==dev_packed) {
         if (gpu_nbor) nbor_k=j+nbor_pitch;
@@ -827,7 +827,7 @@ __kernel void k_vashishta_three_end_vatom(const __global numtyp4 *restrict x_,
           param_costheta_ijk=param5_ijkparam.y;
           param_bigb_ijk=param5_ijkparam.z;
           param_big2b_ijk=param5_ijkparam.w;
-          
+
           numtyp fjx, fjy, fjz, fkx, fky, fkz;
           threebody(delr1x,delr1y,delr1z,eflag,energy);