public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Frederic Weisbecker <fweisbec@gmail.com>
To: Joerg Roedel <joerg.roedel@amd.com>
Cc: mingo@redhat.com, linux-kernel@vger.kernel.org,
	iommu@lists.linux-foundation.org
Subject: Re: [PATCH 03/18] dma-debug: add hash functions for dma_debug_entries
Date: Fri, 6 Mar 2009 14:50:52 +0100	[thread overview]
Message-ID: <20090306135052.GE5988@nowhere> (raw)
In-Reply-To: <1236346229-6618-4-git-send-email-joerg.roedel@amd.com>

On Fri, Mar 06, 2009 at 02:30:14PM +0100, Joerg Roedel wrote:
> Impact: implement necessary functions for the core hash
> 
> Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
> ---
>  lib/dma-debug.c |  101 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
>  1 files changed, 101 insertions(+), 0 deletions(-)
> 
> diff --git a/lib/dma-debug.c b/lib/dma-debug.c
> index 3109971..5ff7d2e 100644
> --- a/lib/dma-debug.c
> +++ b/lib/dma-debug.c
> @@ -18,9 +18,14 @@
>   */
>  
>  #include <linux/dma-debug.h>
> +#include <linux/spinlock.h>
>  #include <linux/types.h>
>  #include <linux/list.h>
>  
> +#define HASH_SIZE       1024ULL
> +#define HASH_FN_SHIFT   13
> +#define HASH_FN_MASK    (HASH_SIZE - 1)
> +
>  enum {
>  	dma_debug_single,
>  	dma_debug_page,
> @@ -40,3 +45,99 @@ struct dma_debug_entry {
>  	int		 sg_mapped_ents;
>  };
>  
> +struct hash_bucket {
> +	struct list_head list;
> +	spinlock_t lock;
> +} __cacheline_aligned_in_smp;
> +
> +/* Hash list to save the allocated dma addresses */
> +static struct hash_bucket dma_entry_hash[HASH_SIZE];
> +
> +/*
> + * Hash related functions
> + *
> + * Every DMA-API request is saved into a struct dma_debug_entry. To
> + * have quick access to these structs they are stored into a hash.
> + */
> +static int hash_fn(struct dma_debug_entry *entry)
> +{
> +	/*
> +	 * Hash function is based on the dma address.
> +	 * We use bits 20-27 here as the index into the hash
> +	 */
> +	return (entry->dev_addr >> HASH_FN_SHIFT) & HASH_FN_MASK;
> +}
> +
> +/*
> + * Request exclusive access to a hash bucket for a given dma_debug_entry.
> + */
> +static struct hash_bucket *get_hash_bucket(struct dma_debug_entry *entry,
> +					   unsigned long *flags)
> +{
> +	int idx = hash_fn(entry);
> +	unsigned long __flags;
> +
> +	spin_lock_irqsave(&dma_entry_hash[idx].lock, __flags);
> +	*flags = __flags;
> +	return &dma_entry_hash[idx];
> +}
> +
> +/*
> + * Give up exclusive access to the hash bucket
> + */
> +static void put_hash_bucket(struct hash_bucket *bucket,
> +			    unsigned long *flags)
> +{
> +	unsigned long __flags = *flags;
> +
> +	spin_unlock_irqrestore(&bucket->lock, __flags);
> +}
> +
> +/*
> + * Search a given entry in the hash bucket list
> + */
> +static struct dma_debug_entry *hash_bucket_find(struct hash_bucket *bucket,
> +						struct dma_debug_entry *ref)
> +{
> +	struct dma_debug_entry *entry;
> +
> +	list_for_each_entry(entry, &bucket->list, list) {
> +		if ((entry->dev_addr == ref->dev_addr) &&
> +		    (entry->dev == ref->dev))
> +			return entry;
> +	}
> +
> +	return NULL;
> +}
> +
> +/*
> + * Add an entry to a hash bucket
> + */
> +static void hash_bucket_add(struct hash_bucket *bucket,
> +			    struct dma_debug_entry *entry)
> +{
> +	list_add_tail(&entry->list, &bucket->list);
> +}
> +
> +/*
> + * Remove entry from a hash bucket list
> + */
> +static void hash_bucket_del(struct dma_debug_entry *entry)
> +{
> +	list_del(&entry->list);
> +}


Perhaps the two wrappers above are unnecessary, since they are actually
used once and only wrap a single list operation. No?

Frederic.

> +/*
> + * Wrapper function for adding an entry to the hash.
> + * This function takes care of locking itself.
> + */
> +static void add_dma_entry(struct dma_debug_entry *entry)
> +{
> +	struct hash_bucket *bucket;
> +	unsigned long flags;
> +
> +	bucket = get_hash_bucket(entry, &flags);
> +	hash_bucket_add(bucket, entry);
> +	put_hash_bucket(bucket, &flags);
> +}
> +
> -- 
> 1.5.6.4
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/


  reply	other threads:[~2009-03-06 13:51 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-06 13:30 [PATCH 0/18] DMA-API debugging facility v4 Joerg Roedel
2009-03-06 13:30 ` [PATCH 01/18] dma-debug: add Kconfig entry Joerg Roedel
2009-03-06 13:30 ` [PATCH 02/18] dma-debug: add header file and core data structures Joerg Roedel
2009-03-06 13:30 ` [PATCH 03/18] dma-debug: add hash functions for dma_debug_entries Joerg Roedel
2009-03-06 13:50   ` Frederic Weisbecker [this message]
2009-03-06 18:45     ` Cyrill Gorcunov
2009-03-06 19:11       ` Frederic Weisbecker
2009-03-06 19:16         ` Cyrill Gorcunov
2009-03-06 19:18           ` Frederic Weisbecker
2009-03-06 19:25           ` Joerg Roedel
2009-03-06 19:38             ` Cyrill Gorcunov
2009-03-06 19:54               ` Joerg Roedel
2009-03-06 13:30 ` [PATCH 04/18] dma-debug: add allocator code Joerg Roedel
2009-03-06 13:30 ` [PATCH 05/18] dma-debug: add initialization code Joerg Roedel
2009-03-06 13:30 ` [PATCH 06/18] dma-debug: add kernel command line parameters Joerg Roedel
2009-03-06 13:30 ` [PATCH 07/18] dma-debug: add debugfs interface Joerg Roedel
2009-03-06 13:45   ` Frederic Weisbecker
2009-03-06 13:30 ` [PATCH 08/18] dma-debug: add core checking functions Joerg Roedel
2009-03-06 13:30 ` [PATCH 09/18] dma-debug: add checking for map/unmap_page/single Joerg Roedel
2009-03-19  1:39   ` FUJITA Tomonori
2009-03-20  8:46     ` Joerg Roedel
2009-03-06 13:30 ` [PATCH 10/18] dma-debug: add add checking for map/unmap_sg Joerg Roedel
2009-03-06 13:30 ` [PATCH 11/18] dma-debug: add checking for [alloc|free]_coherent Joerg Roedel
2009-03-06 13:30 ` [PATCH 12/18] dma-debug: add checks for sync_single_* Joerg Roedel
2009-03-06 13:30 ` [PATCH 13/18] dma-debug: add checks for sync_single_range_* Joerg Roedel
2009-03-06 13:30 ` [PATCH 14/18] dma-debug: add checks for sync_single_sg_* Joerg Roedel
2009-03-06 13:30 ` [PATCH 15/18] dma-debug: add function to dump dma mappings Joerg Roedel
2009-03-06 13:30 ` [PATCH 16/18] dma-debug: x86 architecture bindings Joerg Roedel
2009-03-06 13:30 ` [PATCH 17/18] dma-debug: Documentation update Joerg Roedel
2009-03-06 13:30 ` [PATCH 18/18] dma-debug: print stacktrace of mapping path on unmap error Joerg Roedel

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20090306135052.GE5988@nowhere \
    --to=fweisbec@gmail.com \
    --cc=iommu@lists.linux-foundation.org \
    --cc=joerg.roedel@amd.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox