Clarifying PersistentByteMap (was: Re: [patch] Fix PR java/22189)
Robin Green
greenrd@greenrd.org
Mon Jun 27 12:51:00 GMT 2005
On Mon, Jun 27, 2005 at 10:40:15AM +0100, Andrew Haley wrote:
> Robin Green writes:
> > Hi,
> >
> > Below is a copy of my fix and cleanup for
> > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22189
> >
> > The actual fix is only one line, but I attempted to rename
> > stuff to clarify the code. Please see the bug report for explanation.
> >
> > Needs testing and applying.
>
> Seems reasonable, but:
>
> - Mixes substantive changes and reformatting.
OK, I've posted the bug fix separately. This email deals with the field rename.
The purpose of renaming the capacity field to internal_capacity is
to clarify (for the benefit of newcomers reading the code) that there
are two measures of capacity with different values:
1. The "external_capacity", which is what clients of the PersistentByteMap
class see: the maximum number of elements that are allowed to be inserted.
The bug was that this external capacity changed from 1 to 0 due to a
rounding error.
2. The "internal_capacity", which is the actual number of slots in the
table, and is at least (external_capacity*3)/2.
It could be clearer, I know.
> - Replaces correctly formatted code with differently formatted code
> for no apparent reason;
I probably had mismatching tab/indent settings in Eclipse - hopefully this time
indentations have not been changed.
> at one point a pair of braces is removed
> and replaced by double-spaced blank lines.
I've reverted that unnecessary change.
> - A new local variable, realCapacity, is introduced, but it doesn't
> seem to do anything useful.
I've reverted that unnecessary change. I was trying to preserve the
optimisation of using a local variable rather than reading from an instance
variable several times (I've changed it to read from internal_capacity because
it is dealing with the internal capacity there, not the external capacity).
--
Robin
-------------- next part --------------
--- gnu/gcj/runtime/PersistentByteMap.java~ 2005-06-27 12:38:18.000000000 +0100
+++ gnu/gcj/runtime/PersistentByteMap.java 2005-06-27 12:49:35.000000000 +0100
@@ -67,7 +67,7 @@
static private final int MAGIC = 0;
static private final int VERSION = 4;
- static private final int CAPACITY = 8;
+ static private final int INTERNAL_CAPACITY = 8;
static private final int TABLE_BASE = 12;
static private final int STRING_BASE = 16;
static private final int STRING_SIZE = 20;
@@ -78,7 +78,7 @@
static private final int TABLE_ENTRY_SIZE = 2 * INT_SIZE;
- private int capacity; // real number of entries
+ private int internal_capacity; // number of entries
private int table_base; // offset from start of file, in bytes
private int string_base; // offset from start of file, in bytes
private int string_size; // size of string table, in bytes
@@ -155,7 +155,7 @@
throw new IllegalArgumentException(f.getName());
table_base = getWord (TABLE_BASE);
- capacity = getWord (CAPACITY);
+ internal_capacity = getWord (INTERNAL_CAPACITY);
string_base = getWord (STRING_BASE);
string_size = getWord (STRING_SIZE);
file_size = getWord (FILE_SIZE);
@@ -187,11 +187,11 @@
while (! size.isProbablePrime(10))
size = size.add(two);
- this.capacity = capacity = size.intValue();
+ this.internal_capacity = size.intValue();
}
table_base = 64;
- string_base = table_base + capacity * TABLE_ENTRY_SIZE;
+ string_base = table_base + internal_capacity * TABLE_ENTRY_SIZE;
string_size = 0;
file_size = string_base;
elements = 0;
@@ -207,12 +207,12 @@
fc = raf.getChannel();
buf = fc.map(FileChannel.MapMode.READ_WRITE, 0, raf.length());
- for (int i = 0; i < capacity; i++)
+ for (int i = 0; i < internal_capacity; i++)
putKeyPos(UNUSED_ENTRY, i);
putWord(0x67636a64, MAGIC);
putWord(0x01, VERSION);
- putWord(capacity, CAPACITY);
+ putWord(internal_capacity, INTERNAL_CAPACITY);
putWord(table_base, TABLE_BASE);
putWord(string_base, STRING_BASE);
putWord(file_size, FILE_SIZE);
@@ -305,7 +305,7 @@
+ ((b[1]&0xffL)<<8)
+ ((b[2]&0xffL)<<16)
+ ((b[3]&0xffL)<<24));
- long result = hashIndex % (long)capacity;
+ long result = hashIndex % (long)internal_capacity;
return (int)result;
}
@@ -326,7 +326,7 @@
// not be theoretically as good as open addressing, but it has
// good cache behviour.
hashIndex++;
- hashIndex %= capacity;
+ hashIndex %= internal_capacity;
}
while (true);
}
@@ -360,7 +360,7 @@
}
hashIndex++;
- hashIndex %= capacity;
+ hashIndex %= internal_capacity;
}
while (true);
}
@@ -376,7 +376,7 @@
{
values = new HashMap();
- for (int i = 0; i < capacity; i++)
+ for (int i = 0; i < internal_capacity; i++)
if (getKeyPos(i) != UNUSED_ENTRY)
{
int pos = getValuePos(i);
@@ -437,7 +437,7 @@
{
// With the the table 2/3 full there will be on average 2 probes
// for a successful search and 5 probes for an unsuccessful one.
- return capacity * 2/3;
+ return internal_capacity * 2/3;
}
public void force()
@@ -463,7 +463,7 @@
throws IllegalAccessException
{
// We can use a fast copy if the size of a map has not changed.
- if (this.elements == 0 && t.capacity == this.capacity
+ if (this.elements == 0 && t.internal_capacity == this.internal_capacity
&& t.length == this.length)
{
this.buf.position(0);
@@ -528,7 +528,7 @@
public Object next()
{
count--;
- for (int i = idx; i < capacity; i++)
+ for (int i = idx; i < internal_capacity; i++)
if (getKeyPos(i) != UNUSED_ENTRY)
{
idx = i+1;
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/java-patches/attachments/20050627/781256ae/attachment.sig>
More information about the Java-patches
mailing list