Merge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Wed, 4 Aug 2010 22:23:14 +0000 (15:23 -0700)]
* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (39 commits)
  random: Reorder struct entropy_store to remove padding on 64bits
  padata: update API documentation
  padata: Remove padata_get_cpumask
  crypto: pcrypt - Update pcrypt cpumask according to the padata cpumask notifier
  crypto: pcrypt - Rename pcrypt_instance
  padata: Pass the padata cpumasks to the cpumask_change_notifier chain
  padata: Rearrange set_cpumask functions
  padata: Rename padata_alloc functions
  crypto: pcrypt - Dont calulate a callback cpu on empty callback cpumask
  padata: Check for valid cpumasks
  padata: Allocate cpumask dependend recources in any case
  padata: Fix cpu index counting
  crypto: geode_aes - Convert pci_table entries to PCI_VDEVICE (if PCI_ANY_ID is used)
  pcrypt: Added sysfs interface to pcrypt
  padata: Added sysfs primitives to padata subsystem
  padata: Make two separate cpumasks
  padata: update documentation
  padata: simplify serialization mechanism
  padata: make padata_do_parallel to return zero on success
  padata: Handle empty padata cpumasks
  ...

1  2 
drivers/crypto/talitos.c
kernel/padata.c

diff --combined drivers/crypto/talitos.c
@@@ -720,7 -720,6 +720,6 @@@ struct talitos_ctx 
  #define TALITOS_MDEU_MAX_CONTEXT_SIZE TALITOS_MDEU_CONTEXT_SIZE_SHA384_SHA512
  
  struct talitos_ahash_req_ctx {
-       u64 count;
        u32 hw_context[TALITOS_MDEU_MAX_CONTEXT_SIZE / sizeof(u32)];
        unsigned int hw_context_size;
        u8 buf[HASH_MAX_BLOCK_SIZE];
        unsigned int first;
        unsigned int last;
        unsigned int to_hash_later;
+       u64 nbuf;
        struct scatterlist bufsl[2];
        struct scatterlist *psrc;
  };
@@@ -1183,14 -1183,10 +1183,14 @@@ static size_t sg_copy_end_to_buffer(str
                                /* Copy part of this segment */
                                ignore = skip - offset;
                                len = miter.length - ignore;
 +                              if (boffset + len > buflen)
 +                                      len = buflen - boffset;
                                memcpy(buf + boffset, miter.addr + ignore, len);
                        } else {
 -                              /* Copy all of this segment */
 +                              /* Copy all of this segment (up to buflen) */
                                len = miter.length;
 +                              if (boffset + len > buflen)
 +                                      len = buflen - boffset;
                                memcpy(buf + boffset, miter.addr, len);
                        }
                        boffset += len;
@@@ -1613,6 -1609,7 +1613,7 @@@ static void ahash_done(struct device *d
        if (!req_ctx->last && req_ctx->to_hash_later) {
                /* Position any partial block for next update/final/finup */
                memcpy(req_ctx->buf, req_ctx->bufnext, req_ctx->to_hash_later);
+               req_ctx->nbuf = req_ctx->to_hash_later;
        }
        common_nonsnoop_hash_unmap(dev, edesc, areq);
  
@@@ -1728,7 -1725,7 +1729,7 @@@ static int ahash_init(struct ahash_requ
        struct talitos_ahash_req_ctx *req_ctx = ahash_request_ctx(areq);
  
        /* Initialize the context */
-       req_ctx->count = 0;
+       req_ctx->nbuf = 0;
        req_ctx->first = 1; /* first indicates h/w must init its context */
        req_ctx->swinit = 0; /* assume h/w init of context */
        req_ctx->hw_context_size =
@@@ -1776,52 -1773,54 +1777,54 @@@ static int ahash_process_req(struct aha
                        crypto_tfm_alg_blocksize(crypto_ahash_tfm(tfm));
        unsigned int nbytes_to_hash;
        unsigned int to_hash_later;
-       unsigned int index;
+       unsigned int nsg;
        int chained;
  
-       index = req_ctx->count & (blocksize - 1);
-       req_ctx->count += nbytes;
-       if (!req_ctx->last && (index + nbytes) < blocksize) {
-               /* Buffer the partial block */
+       if (!req_ctx->last && (nbytes + req_ctx->nbuf <= blocksize)) {
+               /* Buffer up to one whole block */
                sg_copy_to_buffer(areq->src,
                                  sg_count(areq->src, nbytes, &chained),
-                                 req_ctx->buf + index, nbytes);
+                                 req_ctx->buf + req_ctx->nbuf, nbytes);
+               req_ctx->nbuf += nbytes;
                return 0;
        }
  
-       if (index) {
-               /* partial block from previous update; chain it in. */
-               sg_init_table(req_ctx->bufsl, (nbytes) ? 2 : 1);
-               sg_set_buf(req_ctx->bufsl, req_ctx->buf, index);
-               if (nbytes)
-                       scatterwalk_sg_chain(req_ctx->bufsl, 2,
-                                            areq->src);
+       /* At least (blocksize + 1) bytes are available to hash */
+       nbytes_to_hash = nbytes + req_ctx->nbuf;
+       to_hash_later = nbytes_to_hash & (blocksize - 1);
+       if (req_ctx->last)
+               to_hash_later = 0;
+       else if (to_hash_later)
+               /* There is a partial block. Hash the full block(s) now */
+               nbytes_to_hash -= to_hash_later;
+       else {
+               /* Keep one block buffered */
+               nbytes_to_hash -= blocksize;
+               to_hash_later = blocksize;
+       }
+       /* Chain in any previously buffered data */
+       if (req_ctx->nbuf) {
+               nsg = (req_ctx->nbuf < nbytes_to_hash) ? 2 : 1;
+               sg_init_table(req_ctx->bufsl, nsg);
+               sg_set_buf(req_ctx->bufsl, req_ctx->buf, req_ctx->nbuf);
+               if (nsg > 1)
+                       scatterwalk_sg_chain(req_ctx->bufsl, 2, areq->src);
                req_ctx->psrc = req_ctx->bufsl;
-       } else {
+       } else
                req_ctx->psrc = areq->src;
+       if (to_hash_later) {
+               int nents = sg_count(areq->src, nbytes, &chained);
+               sg_copy_end_to_buffer(areq->src, nents,
+                                     req_ctx->bufnext,
+                                     to_hash_later,
+                                     nbytes - to_hash_later);
        }
-       nbytes_to_hash =  index + nbytes;
-       if (!req_ctx->last) {
-               to_hash_later = (nbytes_to_hash & (blocksize - 1));
-               if (to_hash_later) {
-                       int nents;
-                       /* Must copy to_hash_later bytes from the end
-                        * to bufnext (a partial block) for later.
-                        */
-                       nents = sg_count(areq->src, nbytes, &chained);
-                       sg_copy_end_to_buffer(areq->src, nents,
-                                             req_ctx->bufnext,
-                                             to_hash_later,
-                                             nbytes - to_hash_later);
-                       /* Adjust count for what will be hashed now */
-                       nbytes_to_hash -= to_hash_later;
-               }
-               req_ctx->to_hash_later = to_hash_later;
-       }
+       req_ctx->to_hash_later = to_hash_later;
  
-       /* allocate extended descriptor */
+       /* Allocate extended descriptor */
        edesc = ahash_edesc_alloc(areq, nbytes_to_hash);
        if (IS_ERR(edesc))
                return PTR_ERR(edesc);
diff --combined kernel/padata.c
  #include <linux/mutex.h>
  #include <linux/sched.h>
  #include <linux/slab.h>
+ #include <linux/sysfs.h>
  #include <linux/rcupdate.h>
  
- #define MAX_SEQ_NR INT_MAX - NR_CPUS
+ #define MAX_SEQ_NR (INT_MAX - NR_CPUS)
  #define MAX_OBJ_NUM 1000
  
  static int padata_index_to_cpu(struct parallel_data *pd, int cpu_index)
  {
        int cpu, target_cpu;
  
-       target_cpu = cpumask_first(pd->cpumask);
+       target_cpu = cpumask_first(pd->cpumask.pcpu);
        for (cpu = 0; cpu < cpu_index; cpu++)
-               target_cpu = cpumask_next(target_cpu, pd->cpumask);
+               target_cpu = cpumask_next(target_cpu, pd->cpumask.pcpu);
  
        return target_cpu;
  }
@@@ -53,26 -54,27 +54,27 @@@ static int padata_cpu_hash(struct padat
         * Hash the sequence numbers to the cpus by taking
         * seq_nr mod. number of cpus in use.
         */
-       cpu_index =  padata->seq_nr % cpumask_weight(pd->cpumask);
+       cpu_index =  padata->seq_nr % cpumask_weight(pd->cpumask.pcpu);
  
        return padata_index_to_cpu(pd, cpu_index);
  }
  
- static void padata_parallel_worker(struct work_struct *work)
+ static void padata_parallel_worker(struct work_struct *parallel_work)
  {
-       struct padata_queue *queue;
+       struct padata_parallel_queue *pqueue;
        struct parallel_data *pd;
        struct padata_instance *pinst;
        LIST_HEAD(local_list);
  
        local_bh_disable();
-       queue = container_of(work, struct padata_queue, pwork);
-       pd = queue->pd;
+       pqueue = container_of(parallel_work,
+                             struct padata_parallel_queue, work);
+       pd = pqueue->pd;
        pinst = pd->pinst;
  
-       spin_lock(&queue->parallel.lock);
-       list_replace_init(&queue->parallel.list, &local_list);
-       spin_unlock(&queue->parallel.lock);
+       spin_lock(&pqueue->parallel.lock);
+       list_replace_init(&pqueue->parallel.list, &local_list);
+       spin_unlock(&pqueue->parallel.lock);
  
        while (!list_empty(&local_list)) {
                struct padata_priv *padata;
@@@ -94,7 -96,7 +96,7 @@@
   * @pinst: padata instance
   * @padata: object to be parallelized
   * @cb_cpu: cpu the serialization callback function will run on,
-  *          must be in the cpumask of padata.
+  *          must be in the serial cpumask of padata(i.e. cpumask.cbcpu).
   *
   * The parallelization callback function will run with BHs off.
   * Note: Every object which is parallelized by padata_do_parallel
@@@ -104,15 -106,18 +106,18 @@@ int padata_do_parallel(struct padata_in
                       struct padata_priv *padata, int cb_cpu)
  {
        int target_cpu, err;
-       struct padata_queue *queue;
+       struct padata_parallel_queue *queue;
        struct parallel_data *pd;
  
        rcu_read_lock_bh();
  
        pd = rcu_dereference(pinst->pd);
  
-       err = 0;
-       if (!(pinst->flags & PADATA_INIT))
+       err = -EINVAL;
+       if (!(pinst->flags & PADATA_INIT) || pinst->flags & PADATA_INVALID)
+               goto out;
+       if (!cpumask_test_cpu(cb_cpu, pd->cpumask.cbcpu))
                goto out;
  
        err =  -EBUSY;
        if (atomic_read(&pd->refcnt) >= MAX_OBJ_NUM)
                goto out;
  
-       err = -EINVAL;
-       if (!cpumask_test_cpu(cb_cpu, pd->cpumask))
-               goto out;
-       err = -EINPROGRESS;
+       err = 0;
        atomic_inc(&pd->refcnt);
        padata->pd = pd;
        padata->cb_cpu = cb_cpu;
        padata->seq_nr = atomic_inc_return(&pd->seq_nr);
  
        target_cpu = padata_cpu_hash(padata);
-       queue = per_cpu_ptr(pd->queue, target_cpu);
+       queue = per_cpu_ptr(pd->pqueue, target_cpu);
  
        spin_lock(&queue->parallel.lock);
        list_add_tail(&padata->list, &queue->parallel.list);
        spin_unlock(&queue->parallel.lock);
  
-       queue_work_on(target_cpu, pinst->wq, &queue->pwork);
+       queue_work_on(target_cpu, pinst->wq, &queue->work);
  
  out:
        rcu_read_unlock_bh();
@@@ -171,84 -172,52 +172,52 @@@ EXPORT_SYMBOL(padata_do_parallel)
   */
  static struct padata_priv *padata_get_next(struct parallel_data *pd)
  {
-       int cpu, num_cpus, empty, calc_seq_nr;
-       int seq_nr, next_nr, overrun, next_overrun;
-       struct padata_queue *queue, *next_queue;
+       int cpu, num_cpus;
+       int next_nr, next_index;
+       struct padata_parallel_queue *queue, *next_queue;
        struct padata_priv *padata;
        struct padata_list *reorder;
  
-       empty = 0;
-       next_nr = -1;
-       next_overrun = 0;
-       next_queue = NULL;
-       num_cpus = cpumask_weight(pd->cpumask);
-       for_each_cpu(cpu, pd->cpumask) {
-               queue = per_cpu_ptr(pd->queue, cpu);
-               reorder = &queue->reorder;
-               /*
-                * Calculate the seq_nr of the object that should be
-                * next in this reorder queue.
-                */
-               overrun = 0;
-               calc_seq_nr = (atomic_read(&queue->num_obj) * num_cpus)
-                              + queue->cpu_index;
+       num_cpus = cpumask_weight(pd->cpumask.pcpu);
  
-               if (unlikely(calc_seq_nr > pd->max_seq_nr)) {
-                       calc_seq_nr = calc_seq_nr - pd->max_seq_nr - 1;
-                       overrun = 1;
-               }
-               if (!list_empty(&reorder->list)) {
-                       padata = list_entry(reorder->list.next,
-                                           struct padata_priv, list);
-                       seq_nr  = padata->seq_nr;
-                       BUG_ON(calc_seq_nr != seq_nr);
-               } else {
-                       seq_nr = calc_seq_nr;
-                       empty++;
-               }
-               if (next_nr < 0 || seq_nr < next_nr
-                   || (next_overrun && !overrun)) {
-                       next_nr = seq_nr;
-                       next_overrun = overrun;
-                       next_queue = queue;
-               }
+       /*
+        * Calculate the percpu reorder queue and the sequence
+        * number of the next object.
+        */
+       next_nr = pd->processed;
+       next_index = next_nr % num_cpus;
+       cpu = padata_index_to_cpu(pd, next_index);
+       next_queue = per_cpu_ptr(pd->pqueue, cpu);
+       if (unlikely(next_nr > pd->max_seq_nr)) {
+               next_nr = next_nr - pd->max_seq_nr - 1;
+               next_index = next_nr % num_cpus;
+               cpu = padata_index_to_cpu(pd, next_index);
+               next_queue = per_cpu_ptr(pd->pqueue, cpu);
+               pd->processed = 0;
        }
  
        padata = NULL;
  
-       if (empty == num_cpus)
-               goto out;
        reorder = &next_queue->reorder;
  
        if (!list_empty(&reorder->list)) {
                padata = list_entry(reorder->list.next,
                                    struct padata_priv, list);
  
-               if (unlikely(next_overrun)) {
-                       for_each_cpu(cpu, pd->cpumask) {
-                               queue = per_cpu_ptr(pd->queue, cpu);
-                               atomic_set(&queue->num_obj, 0);
-                       }
-               }
+               BUG_ON(next_nr != padata->seq_nr);
  
                spin_lock(&reorder->lock);
                list_del_init(&padata->list);
                atomic_dec(&pd->reorder_objects);
                spin_unlock(&reorder->lock);
  
-               atomic_inc(&next_queue->num_obj);
+               pd->processed++;
  
                goto out;
        }
  
-       queue = per_cpu_ptr(pd->queue, smp_processor_id());
+       queue = per_cpu_ptr(pd->pqueue, smp_processor_id());
        if (queue->cpu_index == next_queue->cpu_index) {
                padata = ERR_PTR(-ENODATA);
                goto out;
@@@ -262,7 -231,7 +231,7 @@@ out
  static void padata_reorder(struct parallel_data *pd)
  {
        struct padata_priv *padata;
-       struct padata_queue *queue;
+       struct padata_serial_queue *squeue;
        struct padata_instance *pinst = pd->pinst;
  
        /*
                        return;
                }
  
-               queue = per_cpu_ptr(pd->queue, padata->cb_cpu);
+               squeue = per_cpu_ptr(pd->squeue, padata->cb_cpu);
  
-               spin_lock(&queue->serial.lock);
-               list_add_tail(&padata->list, &queue->serial.list);
-               spin_unlock(&queue->serial.lock);
+               spin_lock(&squeue->serial.lock);
+               list_add_tail(&padata->list, &squeue->serial.list);
+               spin_unlock(&squeue->serial.lock);
  
-               queue_work_on(padata->cb_cpu, pinst->wq, &queue->swork);
+               queue_work_on(padata->cb_cpu, pinst->wq, &squeue->work);
        }
  
        spin_unlock_bh(&pd->lock);
@@@ -333,19 -302,19 +302,19 @@@ static void padata_reorder_timer(unsign
        padata_reorder(pd);
  }
  
- static void padata_serial_worker(struct work_struct *work)
+ static void padata_serial_worker(struct work_struct *serial_work)
  {
-       struct padata_queue *queue;
+       struct padata_serial_queue *squeue;
        struct parallel_data *pd;
        LIST_HEAD(local_list);
  
        local_bh_disable();
-       queue = container_of(work, struct padata_queue, swork);
-       pd = queue->pd;
+       squeue = container_of(serial_work, struct padata_serial_queue, work);
+       pd = squeue->pd;
  
-       spin_lock(&queue->serial.lock);
-       list_replace_init(&queue->serial.list, &local_list);
-       spin_unlock(&queue->serial.lock);
+       spin_lock(&squeue->serial.lock);
+       list_replace_init(&squeue->serial.list, &local_list);
+       spin_unlock(&squeue->serial.lock);
  
        while (!list_empty(&local_list)) {
                struct padata_priv *padata;
  void padata_do_serial(struct padata_priv *padata)
  {
        int cpu;
-       struct padata_queue *queue;
+       struct padata_parallel_queue *pqueue;
        struct parallel_data *pd;
  
        pd = padata->pd;
  
        cpu = get_cpu();
-       queue = per_cpu_ptr(pd->queue, cpu);
+       pqueue = per_cpu_ptr(pd->pqueue, cpu);
  
-       spin_lock(&queue->reorder.lock);
+       spin_lock(&pqueue->reorder.lock);
        atomic_inc(&pd->reorder_objects);
-       list_add_tail(&padata->list, &queue->reorder.list);
-       spin_unlock(&queue->reorder.lock);
+       list_add_tail(&padata->list, &pqueue->reorder.list);
+       spin_unlock(&pqueue->reorder.lock);
  
        put_cpu();
  
  }
  EXPORT_SYMBOL(padata_do_serial);
  
- /* Allocate and initialize the internal cpumask dependend resources. */
- static struct parallel_data *padata_alloc_pd(struct padata_instance *pinst,
-                                            const struct cpumask *cpumask)
+ static int padata_setup_cpumasks(struct parallel_data *pd,
+                                const struct cpumask *pcpumask,
+                                const struct cpumask *cbcpumask)
  {
-       int cpu, cpu_index, num_cpus;
-       struct padata_queue *queue;
-       struct parallel_data *pd;
-       cpu_index = 0;
+       if (!alloc_cpumask_var(&pd->cpumask.pcpu, GFP_KERNEL))
+               return -ENOMEM;
  
-       pd = kzalloc(sizeof(struct parallel_data), GFP_KERNEL);
-       if (!pd)
-               goto err;
+       cpumask_and(pd->cpumask.pcpu, pcpumask, cpu_active_mask);
+       if (!alloc_cpumask_var(&pd->cpumask.cbcpu, GFP_KERNEL)) {
+               free_cpumask_var(pd->cpumask.cbcpu);
+               return -ENOMEM;
+       }
  
-       pd->queue = alloc_percpu(struct padata_queue);
-       if (!pd->queue)
-               goto err_free_pd;
+       cpumask_and(pd->cpumask.cbcpu, cbcpumask, cpu_active_mask);
+       return 0;
+ }
  
-       if (!alloc_cpumask_var(&pd->cpumask, GFP_KERNEL))
-               goto err_free_queue;
+ static void __padata_list_init(struct padata_list *pd_list)
+ {
+       INIT_LIST_HEAD(&pd_list->list);
+       spin_lock_init(&pd_list->lock);
+ }
  
-       cpumask_and(pd->cpumask, cpumask, cpu_active_mask);
+ /* Initialize all percpu queues used by serial workers */
+ static void padata_init_squeues(struct parallel_data *pd)
+ {
+       int cpu;
+       struct padata_serial_queue *squeue;
  
-       for_each_cpu(cpu, pd->cpumask) {
-               queue = per_cpu_ptr(pd->queue, cpu);
+       for_each_cpu(cpu, pd->cpumask.cbcpu) {
+               squeue = per_cpu_ptr(pd->squeue, cpu);
+               squeue->pd = pd;
+               __padata_list_init(&squeue->serial);
+               INIT_WORK(&squeue->work, padata_serial_worker);
+       }
+ }
  
-               queue->pd = pd;
+ /* Initialize all percpu queues used by parallel workers */
+ static void padata_init_pqueues(struct parallel_data *pd)
+ {
+       int cpu_index, num_cpus, cpu;
+       struct padata_parallel_queue *pqueue;
  
-               queue->cpu_index = cpu_index;
+       cpu_index = 0;
+       for_each_cpu(cpu, pd->cpumask.pcpu) {
+               pqueue = per_cpu_ptr(pd->pqueue, cpu);
+               pqueue->pd = pd;
+               pqueue->cpu_index = cpu_index;
                cpu_index++;
  
-               INIT_LIST_HEAD(&queue->reorder.list);
-               INIT_LIST_HEAD(&queue->parallel.list);
-               INIT_LIST_HEAD(&queue->serial.list);
-               spin_lock_init(&queue->reorder.lock);
-               spin_lock_init(&queue->parallel.lock);
-               spin_lock_init(&queue->serial.lock);
-               INIT_WORK(&queue->pwork, padata_parallel_worker);
-               INIT_WORK(&queue->swork, padata_serial_worker);
-               atomic_set(&queue->num_obj, 0);
+               __padata_list_init(&pqueue->reorder);
+               __padata_list_init(&pqueue->parallel);
+               INIT_WORK(&pqueue->work, padata_parallel_worker);
+               atomic_set(&pqueue->num_obj, 0);
        }
  
-       num_cpus = cpumask_weight(pd->cpumask);
-       pd->max_seq_nr = (MAX_SEQ_NR / num_cpus) * num_cpus - 1;
+       num_cpus = cpumask_weight(pd->cpumask.pcpu);
+       pd->max_seq_nr = num_cpus ? (MAX_SEQ_NR / num_cpus) * num_cpus - 1 : 0;
+ }
+ /* Allocate and initialize the internal cpumask dependend resources. */
+ static struct parallel_data *padata_alloc_pd(struct padata_instance *pinst,
+                                            const struct cpumask *pcpumask,
+                                            const struct cpumask *cbcpumask)
+ {
+       struct parallel_data *pd;
  
+       pd = kzalloc(sizeof(struct parallel_data), GFP_KERNEL);
+       if (!pd)
+               goto err;
+       pd->pqueue = alloc_percpu(struct padata_parallel_queue);
+       if (!pd->pqueue)
+               goto err_free_pd;
+       pd->squeue = alloc_percpu(struct padata_serial_queue);
+       if (!pd->squeue)
+               goto err_free_pqueue;
+       if (padata_setup_cpumasks(pd, pcpumask, cbcpumask) < 0)
+               goto err_free_squeue;
+       padata_init_pqueues(pd);
+       padata_init_squeues(pd);
        setup_timer(&pd->timer, padata_reorder_timer, (unsigned long)pd);
        atomic_set(&pd->seq_nr, -1);
        atomic_set(&pd->reorder_objects, 0);
  
        return pd;
  
- err_free_queue:
-       free_percpu(pd->queue);
+ err_free_squeue:
+       free_percpu(pd->squeue);
+ err_free_pqueue:
+       free_percpu(pd->pqueue);
  err_free_pd:
        kfree(pd);
  err:
  
  static void padata_free_pd(struct parallel_data *pd)
  {
-       free_cpumask_var(pd->cpumask);
-       free_percpu(pd->queue);
+       free_cpumask_var(pd->cpumask.pcpu);
+       free_cpumask_var(pd->cpumask.cbcpu);
+       free_percpu(pd->pqueue);
+       free_percpu(pd->squeue);
        kfree(pd);
  }
  
  static void padata_flush_queues(struct parallel_data *pd)
  {
        int cpu;
-       struct padata_queue *queue;
+       struct padata_parallel_queue *pqueue;
+       struct padata_serial_queue *squeue;
  
-       for_each_cpu(cpu, pd->cpumask) {
-               queue = per_cpu_ptr(pd->queue, cpu);
-               flush_work(&queue->pwork);
+       for_each_cpu(cpu, pd->cpumask.pcpu) {
+               pqueue = per_cpu_ptr(pd->pqueue, cpu);
+               flush_work(&pqueue->work);
        }
  
        del_timer_sync(&pd->timer);
        if (atomic_read(&pd->reorder_objects))
                padata_reorder(pd);
  
-       for_each_cpu(cpu, pd->cpumask) {
-               queue = per_cpu_ptr(pd->queue, cpu);
-               flush_work(&queue->swork);
+       for_each_cpu(cpu, pd->cpumask.cbcpu) {
+               squeue = per_cpu_ptr(pd->squeue, cpu);
+               flush_work(&squeue->work);
        }
  
        BUG_ON(atomic_read(&pd->refcnt) != 0);
  }
  
+ static void __padata_start(struct padata_instance *pinst)
+ {
+       pinst->flags |= PADATA_INIT;
+ }
+ static void __padata_stop(struct padata_instance *pinst)
+ {
+       if (!(pinst->flags & PADATA_INIT))
+               return;
+       pinst->flags &= ~PADATA_INIT;
+       synchronize_rcu();
+       get_online_cpus();
+       padata_flush_queues(pinst->pd);
+       put_online_cpus();
+ }
  /* Replace the internal control stucture with a new one. */
  static void padata_replace(struct padata_instance *pinst,
                           struct parallel_data *pd_new)
  {
        struct parallel_data *pd_old = pinst->pd;
+       int notification_mask = 0;
  
        pinst->flags |= PADATA_RESET;
  
  
        synchronize_rcu();
  
+       if (!cpumask_equal(pd_old->cpumask.pcpu, pd_new->cpumask.pcpu))
+               notification_mask |= PADATA_CPU_PARALLEL;
+       if (!cpumask_equal(pd_old->cpumask.cbcpu, pd_new->cpumask.cbcpu))
+               notification_mask |= PADATA_CPU_SERIAL;
        padata_flush_queues(pd_old);
        padata_free_pd(pd_old);
  
+       if (notification_mask)
+               blocking_notifier_call_chain(&pinst->cpumask_change_notifier,
+                                            notification_mask,
+                                            &pd_new->cpumask);
        pinst->flags &= ~PADATA_RESET;
  }
  
  /**
-  * padata_set_cpumask - set the cpumask that padata should use
+  * padata_register_cpumask_notifier - Registers a notifier that will be called
+  *                             if either pcpu or cbcpu or both cpumasks change.
   *
-  * @pinst: padata instance
-  * @cpumask: the cpumask to use
+  * @pinst: A poineter to padata instance
+  * @nblock: A pointer to notifier block.
   */
- int padata_set_cpumask(struct padata_instance *pinst,
-                       cpumask_var_t cpumask)
+ int padata_register_cpumask_notifier(struct padata_instance *pinst,
+                                    struct notifier_block *nblock)
  {
+       return blocking_notifier_chain_register(&pinst->cpumask_change_notifier,
+                                               nblock);
+ }
+ EXPORT_SYMBOL(padata_register_cpumask_notifier);
+ /**
+  * padata_unregister_cpumask_notifier - Unregisters cpumask notifier
+  *        registered earlier  using padata_register_cpumask_notifier
+  *
+  * @pinst: A pointer to data instance.
+  * @nlock: A pointer to notifier block.
+  */
+ int padata_unregister_cpumask_notifier(struct padata_instance *pinst,
+                                      struct notifier_block *nblock)
+ {
+       return blocking_notifier_chain_unregister(
+               &pinst->cpumask_change_notifier,
+               nblock);
+ }
+ EXPORT_SYMBOL(padata_unregister_cpumask_notifier);
+ /* If cpumask contains no active cpu, we mark the instance as invalid. */
+ static bool padata_validate_cpumask(struct padata_instance *pinst,
+                                   const struct cpumask *cpumask)
+ {
+       if (!cpumask_intersects(cpumask, cpu_active_mask)) {
+               pinst->flags |= PADATA_INVALID;
+               return false;
+       }
+       pinst->flags &= ~PADATA_INVALID;
+       return true;
+ }
+ static int __padata_set_cpumasks(struct padata_instance *pinst,
+                                cpumask_var_t pcpumask,
+                                cpumask_var_t cbcpumask)
+ {
+       int valid;
        struct parallel_data *pd;
-       int err = 0;
+       valid = padata_validate_cpumask(pinst, pcpumask);
+       if (!valid) {
+               __padata_stop(pinst);
+               goto out_replace;
+       }
+       valid = padata_validate_cpumask(pinst, cbcpumask);
+       if (!valid)
+               __padata_stop(pinst);
+ out_replace:
+       pd = padata_alloc_pd(pinst, pcpumask, cbcpumask);
+       if (!pd)
+               return -ENOMEM;
+       cpumask_copy(pinst->cpumask.pcpu, pcpumask);
+       cpumask_copy(pinst->cpumask.cbcpu, cbcpumask);
+       padata_replace(pinst, pd);
+       if (valid)
+               __padata_start(pinst);
+       return 0;
+ }
+ /**
+  * padata_set_cpumasks - Set both parallel and serial cpumasks. The first
+  *                       one is used by parallel workers and the second one
+  *                       by the wokers doing serialization.
+  *
+  * @pinst: padata instance
+  * @pcpumask: the cpumask to use for parallel workers
+  * @cbcpumask: the cpumsak to use for serial workers
+  */
+ int padata_set_cpumasks(struct padata_instance *pinst, cpumask_var_t pcpumask,
+                       cpumask_var_t cbcpumask)
+ {
+       int err;
  
        mutex_lock(&pinst->lock);
+       get_online_cpus();
  
+       err = __padata_set_cpumasks(pinst, pcpumask, cbcpumask);
+       put_online_cpus();
+       mutex_unlock(&pinst->lock);
+       return err;
+ }
+ EXPORT_SYMBOL(padata_set_cpumasks);
+ /**
+  * padata_set_cpumask: Sets specified by @cpumask_type cpumask to the value
+  *                     equivalent to @cpumask.
+  *
+  * @pinst: padata instance
+  * @cpumask_type: PADATA_CPU_SERIAL or PADATA_CPU_PARALLEL corresponding
+  *                to parallel and serial cpumasks respectively.
+  * @cpumask: the cpumask to use
+  */
+ int padata_set_cpumask(struct padata_instance *pinst, int cpumask_type,
+                      cpumask_var_t cpumask)
+ {
+       struct cpumask *serial_mask, *parallel_mask;
+       int err = -EINVAL;
+       mutex_lock(&pinst->lock);
        get_online_cpus();
  
-       pd = padata_alloc_pd(pinst, cpumask);
-       if (!pd) {
-               err = -ENOMEM;
-               goto out;
+       switch (cpumask_type) {
+       case PADATA_CPU_PARALLEL:
+               serial_mask = pinst->cpumask.cbcpu;
+               parallel_mask = cpumask;
+               break;
+       case PADATA_CPU_SERIAL:
+               parallel_mask = pinst->cpumask.pcpu;
+               serial_mask = cpumask;
+               break;
+       default:
+                goto out;
        }
  
-       cpumask_copy(pinst->cpumask, cpumask);
-       padata_replace(pinst, pd);
+       err =  __padata_set_cpumasks(pinst, parallel_mask, serial_mask);
  
  out:
        put_online_cpus();
        mutex_unlock(&pinst->lock);
  
        return err;
@@@ -543,30 -695,48 +695,48 @@@ static int __padata_add_cpu(struct pada
        struct parallel_data *pd;
  
        if (cpumask_test_cpu(cpu, cpu_active_mask)) {
-               pd = padata_alloc_pd(pinst, pinst->cpumask);
+               pd = padata_alloc_pd(pinst, pinst->cpumask.pcpu,
+                                    pinst->cpumask.cbcpu);
                if (!pd)
                        return -ENOMEM;
  
                padata_replace(pinst, pd);
+               if (padata_validate_cpumask(pinst, pinst->cpumask.pcpu) &&
+                   padata_validate_cpumask(pinst, pinst->cpumask.cbcpu))
+                       __padata_start(pinst);
        }
  
        return 0;
  }
  
- /**
-  * padata_add_cpu - add a cpu to the padata cpumask
+  /**
+  * padata_add_cpu - add a cpu to one or both(parallel and serial)
+  *                  padata cpumasks.
   *
   * @pinst: padata instance
   * @cpu: cpu to add
+  * @mask: bitmask of flags specifying to which cpumask @cpu shuld be added.
+  *        The @mask may be any combination of the following flags:
+  *          PADATA_CPU_SERIAL   - serial cpumask
+  *          PADATA_CPU_PARALLEL - parallel cpumask
   */
- int padata_add_cpu(struct padata_instance *pinst, int cpu)
+ int padata_add_cpu(struct padata_instance *pinst, int cpu, int mask)
  {
        int err;
  
+       if (!(mask & (PADATA_CPU_SERIAL | PADATA_CPU_PARALLEL)))
+               return -EINVAL;
        mutex_lock(&pinst->lock);
  
        get_online_cpus();
-       cpumask_set_cpu(cpu, pinst->cpumask);
+       if (mask & PADATA_CPU_SERIAL)
+               cpumask_set_cpu(cpu, pinst->cpumask.cbcpu);
+       if (mask & PADATA_CPU_PARALLEL)
+               cpumask_set_cpu(cpu, pinst->cpumask.pcpu);
        err = __padata_add_cpu(pinst, cpu);
        put_online_cpus();
  
@@@ -578,10 -748,16 +748,16 @@@ EXPORT_SYMBOL(padata_add_cpu)
  
  static int __padata_remove_cpu(struct padata_instance *pinst, int cpu)
  {
-       struct parallel_data *pd;
+       struct parallel_data *pd = NULL;
  
        if (cpumask_test_cpu(cpu, cpu_online_mask)) {
-               pd = padata_alloc_pd(pinst, pinst->cpumask);
+               if (!padata_validate_cpumask(pinst, pinst->cpumask.pcpu) ||
+                   !padata_validate_cpumask(pinst, pinst->cpumask.cbcpu))
+                       __padata_stop(pinst);
+               pd = padata_alloc_pd(pinst, pinst->cpumask.pcpu,
+                                    pinst->cpumask.cbcpu);
                if (!pd)
                        return -ENOMEM;
  
        return 0;
  }
  
- /**
-  * padata_remove_cpu - remove a cpu from the padata cpumask
+  /**
+  * padata_remove_cpu - remove a cpu from the one or both(serial and paralell)
+  *                     padata cpumasks.
   *
   * @pinst: padata instance
   * @cpu: cpu to remove
+  * @mask: bitmask specifying from which cpumask @cpu should be removed
+  *        The @mask may be any combination of the following flags:
+  *          PADATA_CPU_SERIAL   - serial cpumask
+  *          PADATA_CPU_PARALLEL - parallel cpumask
   */
- int padata_remove_cpu(struct padata_instance *pinst, int cpu)
+ int padata_remove_cpu(struct padata_instance *pinst, int cpu, int mask)
  {
        int err;
  
+       if (!(mask & (PADATA_CPU_SERIAL | PADATA_CPU_PARALLEL)))
+               return -EINVAL;
        mutex_lock(&pinst->lock);
  
        get_online_cpus();
-       cpumask_clear_cpu(cpu, pinst->cpumask);
+       if (mask & PADATA_CPU_SERIAL)
+               cpumask_clear_cpu(cpu, pinst->cpumask.cbcpu);
+       if (mask & PADATA_CPU_PARALLEL)
+               cpumask_clear_cpu(cpu, pinst->cpumask.pcpu);
        err = __padata_remove_cpu(pinst, cpu);
        put_online_cpus();
  
@@@ -619,11 -807,20 +807,20 @@@ EXPORT_SYMBOL(padata_remove_cpu)
   *
   * @pinst: padata instance to start
   */
- void padata_start(struct padata_instance *pinst)
+ int padata_start(struct padata_instance *pinst)
  {
+       int err = 0;
        mutex_lock(&pinst->lock);
-       pinst->flags |= PADATA_INIT;
+       if (pinst->flags & PADATA_INVALID)
+               err =-EINVAL;
+        __padata_start(pinst);
        mutex_unlock(&pinst->lock);
+       return err;
  }
  EXPORT_SYMBOL(padata_start);
  
  void padata_stop(struct padata_instance *pinst)
  {
        mutex_lock(&pinst->lock);
-       pinst->flags &= ~PADATA_INIT;
+       __padata_stop(pinst);
        mutex_unlock(&pinst->lock);
  }
  EXPORT_SYMBOL(padata_stop);
  
  #ifdef CONFIG_HOTPLUG_CPU
+ static inline int pinst_has_cpu(struct padata_instance *pinst, int cpu)
+ {
+       return cpumask_test_cpu(cpu, pinst->cpumask.pcpu) ||
+               cpumask_test_cpu(cpu, pinst->cpumask.cbcpu);
+ }
  static int padata_cpu_callback(struct notifier_block *nfb,
                               unsigned long action, void *hcpu)
  {
        switch (action) {
        case CPU_ONLINE:
        case CPU_ONLINE_FROZEN:
-               if (!cpumask_test_cpu(cpu, pinst->cpumask))
+               if (!pinst_has_cpu(pinst, cpu))
                        break;
                mutex_lock(&pinst->lock);
                err = __padata_add_cpu(pinst, cpu);
                mutex_unlock(&pinst->lock);
                if (err)
 -                      return NOTIFY_BAD;
 +                      return notifier_from_errno(err);
                break;
  
        case CPU_DOWN_PREPARE:
        case CPU_DOWN_PREPARE_FROZEN:
-               if (!cpumask_test_cpu(cpu, pinst->cpumask))
+               if (!pinst_has_cpu(pinst, cpu))
                        break;
                mutex_lock(&pinst->lock);
                err = __padata_remove_cpu(pinst, cpu);
                mutex_unlock(&pinst->lock);
                if (err)
 -                      return NOTIFY_BAD;
 +                      return notifier_from_errno(err);
                break;
  
        case CPU_UP_CANCELED:
        case CPU_UP_CANCELED_FROZEN:
-               if (!cpumask_test_cpu(cpu, pinst->cpumask))
+               if (!pinst_has_cpu(pinst, cpu))
                        break;
                mutex_lock(&pinst->lock);
                __padata_remove_cpu(pinst, cpu);
  
        case CPU_DOWN_FAILED:
        case CPU_DOWN_FAILED_FROZEN:
-               if (!cpumask_test_cpu(cpu, pinst->cpumask))
+               if (!pinst_has_cpu(pinst, cpu))
                        break;
                mutex_lock(&pinst->lock);
                __padata_add_cpu(pinst, cpu);
  }
  #endif
  
+ static void __padata_free(struct padata_instance *pinst)
+ {
+ #ifdef CONFIG_HOTPLUG_CPU
+       unregister_hotcpu_notifier(&pinst->cpu_notifier);
+ #endif
+       padata_stop(pinst);
+       padata_free_pd(pinst->pd);
+       free_cpumask_var(pinst->cpumask.pcpu);
+       free_cpumask_var(pinst->cpumask.cbcpu);
+       kfree(pinst);
+ }
+ #define kobj2pinst(_kobj)                                     \
+       container_of(_kobj, struct padata_instance, kobj)
+ #define attr2pentry(_attr)                                    \
+       container_of(_attr, struct padata_sysfs_entry, attr)
+ static void padata_sysfs_release(struct kobject *kobj)
+ {
+       struct padata_instance *pinst = kobj2pinst(kobj);
+       __padata_free(pinst);
+ }
+ struct padata_sysfs_entry {
+       struct attribute attr;
+       ssize_t (*show)(struct padata_instance *, struct attribute *, char *);
+       ssize_t (*store)(struct padata_instance *, struct attribute *,
+                        const char *, size_t);
+ };
+ static ssize_t show_cpumask(struct padata_instance *pinst,
+                           struct attribute *attr,  char *buf)
+ {
+       struct cpumask *cpumask;
+       ssize_t len;
+       mutex_lock(&pinst->lock);
+       if (!strcmp(attr->name, "serial_cpumask"))
+               cpumask = pinst->cpumask.cbcpu;
+       else
+               cpumask = pinst->cpumask.pcpu;
+       len = bitmap_scnprintf(buf, PAGE_SIZE, cpumask_bits(cpumask),
+                              nr_cpu_ids);
+       if (PAGE_SIZE - len < 2)
+               len = -EINVAL;
+       else
+               len += sprintf(buf + len, "\n");
+       mutex_unlock(&pinst->lock);
+       return len;
+ }
+ static ssize_t store_cpumask(struct padata_instance *pinst,
+                            struct attribute *attr,
+                            const char *buf, size_t count)
+ {
+       cpumask_var_t new_cpumask;
+       ssize_t ret;
+       int mask_type;
+       if (!alloc_cpumask_var(&new_cpumask, GFP_KERNEL))
+               return -ENOMEM;
+       ret = bitmap_parse(buf, count, cpumask_bits(new_cpumask),
+                          nr_cpumask_bits);
+       if (ret < 0)
+               goto out;
+       mask_type = !strcmp(attr->name, "serial_cpumask") ?
+               PADATA_CPU_SERIAL : PADATA_CPU_PARALLEL;
+       ret = padata_set_cpumask(pinst, mask_type, new_cpumask);
+       if (!ret)
+               ret = count;
+ out:
+       free_cpumask_var(new_cpumask);
+       return ret;
+ }
+ #define PADATA_ATTR_RW(_name, _show_name, _store_name)                \
+       static struct padata_sysfs_entry _name##_attr =         \
+               __ATTR(_name, 0644, _show_name, _store_name)
+ #define PADATA_ATTR_RO(_name, _show_name)             \
+       static struct padata_sysfs_entry _name##_attr = \
+               __ATTR(_name, 0400, _show_name, NULL)
+ PADATA_ATTR_RW(serial_cpumask, show_cpumask, store_cpumask);
+ PADATA_ATTR_RW(parallel_cpumask, show_cpumask, store_cpumask);
+ /*
+  * Padata sysfs provides the following objects:
+  * serial_cpumask   [RW] - cpumask for serial workers
+  * parallel_cpumask [RW] - cpumask for parallel workers
+  */
+ static struct attribute *padata_default_attrs[] = {
+       &serial_cpumask_attr.attr,
+       &parallel_cpumask_attr.attr,
+       NULL,
+ };
+ static ssize_t padata_sysfs_show(struct kobject *kobj,
+                                struct attribute *attr, char *buf)
+ {
+       struct padata_instance *pinst;
+       struct padata_sysfs_entry *pentry;
+       ssize_t ret = -EIO;
+       pinst = kobj2pinst(kobj);
+       pentry = attr2pentry(attr);
+       if (pentry->show)
+               ret = pentry->show(pinst, attr, buf);
+       return ret;
+ }
+ static ssize_t padata_sysfs_store(struct kobject *kobj, struct attribute *attr,
+                                 const char *buf, size_t count)
+ {
+       struct padata_instance *pinst;
+       struct padata_sysfs_entry *pentry;
+       ssize_t ret = -EIO;
+       pinst = kobj2pinst(kobj);
+       pentry = attr2pentry(attr);
+       if (pentry->show)
+               ret = pentry->store(pinst, attr, buf, count);
+       return ret;
+ }
+ static const struct sysfs_ops padata_sysfs_ops = {
+       .show = padata_sysfs_show,
+       .store = padata_sysfs_store,
+ };
+ static struct kobj_type padata_attr_type = {
+       .sysfs_ops = &padata_sysfs_ops,
+       .default_attrs = padata_default_attrs,
+       .release = padata_sysfs_release,
+ };
  /**
-  * padata_alloc - allocate and initialize a padata instance
+  * padata_alloc_possible - Allocate and initialize padata instance.
+  *                         Use the cpu_possible_mask for serial and
+  *                         parallel workers.
   *
-  * @cpumask: cpumask that padata uses for parallelization
   * @wq: workqueue to use for the allocated padata instance
   */
- struct padata_instance *padata_alloc(const struct cpumask *cpumask,
-                                    struct workqueue_struct *wq)
+ struct padata_instance *padata_alloc_possible(struct workqueue_struct *wq)
+ {
+       return padata_alloc(wq, cpu_possible_mask, cpu_possible_mask);
+ }
+ EXPORT_SYMBOL(padata_alloc_possible);
+ /**
+  * padata_alloc - allocate and initialize a padata instance and specify
+  *                cpumasks for serial and parallel workers.
+  *
+  * @wq: workqueue to use for the allocated padata instance
+  * @pcpumask: cpumask that will be used for padata parallelization
+  * @cbcpumask: cpumask that will be used for padata serialization
+  */
+ struct padata_instance *padata_alloc(struct workqueue_struct *wq,
+                                    const struct cpumask *pcpumask,
+                                    const struct cpumask *cbcpumask)
  {
        struct padata_instance *pinst;
-       struct parallel_data *pd;
+       struct parallel_data *pd = NULL;
  
        pinst = kzalloc(sizeof(struct padata_instance), GFP_KERNEL);
        if (!pinst)
                goto err;
  
        get_online_cpus();
-       pd = padata_alloc_pd(pinst, cpumask);
-       if (!pd)
+       if (!alloc_cpumask_var(&pinst->cpumask.pcpu, GFP_KERNEL))
                goto err_free_inst;
+       if (!alloc_cpumask_var(&pinst->cpumask.cbcpu, GFP_KERNEL)) {
+               free_cpumask_var(pinst->cpumask.pcpu);
+               goto err_free_inst;
+       }
+       if (!padata_validate_cpumask(pinst, pcpumask) ||
+           !padata_validate_cpumask(pinst, cbcpumask))
+               goto err_free_masks;
  
-       if (!alloc_cpumask_var(&pinst->cpumask, GFP_KERNEL))
-               goto err_free_pd;
+       pd = padata_alloc_pd(pinst, pcpumask, cbcpumask);
+       if (!pd)
+               goto err_free_masks;
  
        rcu_assign_pointer(pinst->pd, pd);
  
        pinst->wq = wq;
  
-       cpumask_copy(pinst->cpumask, cpumask);
+       cpumask_copy(pinst->cpumask.pcpu, pcpumask);
+       cpumask_copy(pinst->cpumask.cbcpu, cbcpumask);
  
        pinst->flags = 0;
  
  
        put_online_cpus();
  
+       BLOCKING_INIT_NOTIFIER_HEAD(&pinst->cpumask_change_notifier);
+       kobject_init(&pinst->kobj, &padata_attr_type);
        mutex_init(&pinst->lock);
  
        return pinst;
  
- err_free_pd:
-       padata_free_pd(pd);
+ err_free_masks:
+       free_cpumask_var(pinst->cpumask.pcpu);
+       free_cpumask_var(pinst->cpumask.cbcpu);
  err_free_inst:
        kfree(pinst);
        put_online_cpus();
@@@ -756,19 -1130,6 +1130,6 @@@ EXPORT_SYMBOL(padata_alloc)
   */
  void padata_free(struct padata_instance *pinst)
  {
-       padata_stop(pinst);
-       synchronize_rcu();
- #ifdef CONFIG_HOTPLUG_CPU
-       unregister_hotcpu_notifier(&pinst->cpu_notifier);
- #endif
-       get_online_cpus();
-       padata_flush_queues(pinst->pd);
-       put_online_cpus();
-       padata_free_pd(pinst->pd);
-       free_cpumask_var(pinst->cpumask);
-       kfree(pinst);
+       kobject_put(&pinst->kobj);
  }
  EXPORT_SYMBOL(padata_free);