You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@directmemory.apache.org by "Benoit Perroud (Updated) (JIRA)" <ji...@apache.org> on 2012/03/02 11:13:54 UTC

[jira] [Updated] (DIRECTMEMORY-49) MemoryManagerService.update does not reuse the same pointer

     [ https://issues.apache.org/jira/browse/DIRECTMEMORY-49?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Benoit Perroud updated DIRECTMEMORY-49:
---------------------------------------

    Description: 
MemoryManagerService.update actually free and store again the payload, which leads to another pointer allocation. This is fine in lots of case, but when the buffer is full, we still want the update to succeed.

Moreover, update should throw an exception (BufferOverFlowException for example) when the new payload data is bigger than the previous.

Considering the following test, where I allocate just enough space for 1 object, I store it, and then want to update it content. I could hope this use case succeed, but it's actually not the case

{code}
@Test
public void testOffHeapMemoryBufferUpdate() {
		
	final int NUMBER_OF_OBJECTS = 1;
	final int BUFFER_SIZE = NUMBER_OF_OBJECTS * 4; // allocate 4 bytes
	
	final OffHeapMemoryBuffer offHeapMemoryBuffer = OffHeapMemoryBufferImpl.createNew(BUFFER_SIZE);

        // generate 4 random bytes, store, read back and assert
	final byte[] payload = generateRandomPayload(4);
	final Pointer pointer = offHeapMemoryBuffer.store(payload);
	Assert.assertNotNull(pointer);
	Assert.assertEquals(new String(payload), new String(offHeapMemoryBuffer.retrieve(pointer)));
	
        // generate another payload, update it, read it and assert.
	byte[] otherPayload = generateRandomPayload(4);
	final Pointer otherPointer = offHeapMemoryBuffer.update(pointer, otherPayload);
	Assert.assertNotNull(otherPointer);
	Assert.assertEquals(new String(otherPayload), new String(offHeapMemoryBuffer.retrieve(otherPointer)));
}
{code}

  was:
OffHeapMemoryBuffer.update actually free and store again the payload, which leads to another pointer allocation. This is fine in lots of case, but when the buffer is full, we still want the update to succeed.

Moreover, update should throw an exception (BufferOverFlowException for example) when the new payload data is bigger than the previous.

Considering the following test, where I allocate just enough space for 1 object, I store it, and then want to update it content. I could hope this use case succeed, but it's actually not the case

{code}
@Test
public void testOffHeapMemoryBufferUpdate() {
		
	final int NUMBER_OF_OBJECTS = 1;
	final int BUFFER_SIZE = NUMBER_OF_OBJECTS * 4; // allocate 4 bytes
	
	final OffHeapMemoryBuffer offHeapMemoryBuffer = OffHeapMemoryBufferImpl.createNew(BUFFER_SIZE);

        // generate 4 random bytes, store, read back and assert
	final byte[] payload = generateRandomPayload(4);
	final Pointer pointer = offHeapMemoryBuffer.store(payload);
	Assert.assertNotNull(pointer);
	Assert.assertEquals(new String(payload), new String(offHeapMemoryBuffer.retrieve(pointer)));
	
        // generate another payload, update it, read it and assert.
	byte[] otherPayload = generateRandomPayload(4);
	final Pointer otherPointer = offHeapMemoryBuffer.update(pointer, otherPayload);
	Assert.assertNotNull(otherPointer);
	Assert.assertEquals(new String(otherPayload), new String(offHeapMemoryBuffer.retrieve(otherPointer)));
}
{code}

        Summary: MemoryManagerService.update does not reuse the same pointer  (was: OffHeapMemoryBuffer.update does not reuse the same pointer)
    
> MemoryManagerService.update does not reuse the same pointer
> -----------------------------------------------------------
>
>                 Key: DIRECTMEMORY-49
>                 URL: https://issues.apache.org/jira/browse/DIRECTMEMORY-49
>             Project: Apache DirectMemory
>          Issue Type: Bug
>            Reporter: Benoit Perroud
>            Priority: Minor
>
> MemoryManagerService.update actually free and store again the payload, which leads to another pointer allocation. This is fine in lots of case, but when the buffer is full, we still want the update to succeed.
> Moreover, update should throw an exception (BufferOverFlowException for example) when the new payload data is bigger than the previous.
> Considering the following test, where I allocate just enough space for 1 object, I store it, and then want to update it content. I could hope this use case succeed, but it's actually not the case
> {code}
> @Test
> public void testOffHeapMemoryBufferUpdate() {
> 		
> 	final int NUMBER_OF_OBJECTS = 1;
> 	final int BUFFER_SIZE = NUMBER_OF_OBJECTS * 4; // allocate 4 bytes
> 	
> 	final OffHeapMemoryBuffer offHeapMemoryBuffer = OffHeapMemoryBufferImpl.createNew(BUFFER_SIZE);
>         // generate 4 random bytes, store, read back and assert
> 	final byte[] payload = generateRandomPayload(4);
> 	final Pointer pointer = offHeapMemoryBuffer.store(payload);
> 	Assert.assertNotNull(pointer);
> 	Assert.assertEquals(new String(payload), new String(offHeapMemoryBuffer.retrieve(pointer)));
> 	
>         // generate another payload, update it, read it and assert.
> 	byte[] otherPayload = generateRandomPayload(4);
> 	final Pointer otherPointer = offHeapMemoryBuffer.update(pointer, otherPayload);
> 	Assert.assertNotNull(otherPointer);
> 	Assert.assertEquals(new String(otherPayload), new String(offHeapMemoryBuffer.retrieve(otherPointer)));
> }
> {code}

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira