search for: try_again

...@@ -2190,7 +2191,7 @@ static int ocfs2_xattr_can_be_in_inode(struct inode *inode, last += 1; } - free = min_offs - ((void *)last - xs->base) - sizeof(__u32); + free = min_offs - ((void *)last - xs->base) - OCFS2_XATTR_HEADER_GAP; if (free < 0) return 0; @@ -5060,8 +5061,8 @@ try_again: xh_free_start = le16_to_cpu(xh->xh_free_start); header_size = sizeof(struct ocfs2_xattr_header) + count * sizeof(struct ocfs2_xattr_entry); - max_free = OCFS2_XATTR_BUCKET_SIZE - - le16_to_cpu(xh->xh_name_value_len) - header_size; + max_free = OCFS2_XATTR_BUCKET_SIZE - header_size -...

...#39;re trying to optimize this write. Why do you need a whole byte for that? Surely a cmpxchg loop with the right atomic op can't be _that_ much slower? Its far more readable and likely avoids that steal fail below as well. > + return 1; > + } else if (old == _QSPINLOCK_LOCKED) { > +try_again: > + /* > + * Wait until the lock byte is cleared to get the lock > + */ > + do { > + cpu_relax(); > + } while (ACCESS_ONCE(qlock->lock)); > + /* > + * Set the lock bit & clear the waiting bit > + */ > + if (cmpxchg(&qlock->lock_wait, _QSPI...

...#39;re trying to optimize this write. Why do you need a whole byte for that? Surely a cmpxchg loop with the right atomic op can't be _that_ much slower? Its far more readable and likely avoids that steal fail below as well. > + return 1; > + } else if (old == _QSPINLOCK_LOCKED) { > +try_again: > + /* > + * Wait until the lock byte is cleared to get the lock > + */ > + do { > + cpu_relax(); > + } while (ACCESS_ONCE(qlock->lock)); > + /* > + * Set the lock bit & clear the waiting bit > + */ > + if (cmpxchg(&qlock->lock_wait, _QSPI...

...ace_info->groups_sem); + goto search; } + } + up_read(&space_info->groups_sem); - block_group = list_entry(cur, struct btrfs_block_group_cache, - list); - atomic_inc(&block_group->count); + if (!ins->objectid && (empty_size || allowed_chunk_alloc)) { + int try_again = empty_size; - search_start = block_group->key.objectid; - cur = cur->next; + total_needed -= empty_size; + empty_size = 0; + + if (allowed_chunk_alloc) { + ret = do_chunk_alloc(trans, root, num_bytes + + 2 * 1024 * 1024, data, 1); + if (!ret) + try_again = 1; + allo...

...SPINLOCK_WAITING))) + return 0; + + old = xchg(&qlock->lock_wait, _QSPINLOCK_WAITING|_QSPINLOCK_LOCKED); + + if (old == 0) { + /* + * Got the lock, can clear the waiting bit now + */ + smp_u8_store_release(&qlock->wait, 0); + return 1; + } else if (old == _QSPINLOCK_LOCKED) { +try_again: + /* + * Wait until the lock byte is cleared to get the lock + */ + do { + cpu_relax(); + } while (ACCESS_ONCE(qlock->lock)); + /* + * Set the lock bit & clear the waiting bit + */ + if (cmpxchg(&qlock->lock_wait, _QSPINLOCK_WAITING, + _QSPINLOCK_LOCKED) == _QSPIN...

...SPINLOCK_WAITING))) + return 0; + + old = xchg(&qlock->lock_wait, _QSPINLOCK_WAITING|_QSPINLOCK_LOCKED); + + if (old == 0) { + /* + * Got the lock, can clear the waiting bit now + */ + smp_u8_store_release(&qlock->wait, 0); + return 1; + } else if (old == _QSPINLOCK_LOCKED) { +try_again: + /* + * Wait until the lock byte is cleared to get the lock + */ + do { + cpu_relax(); + } while (ACCESS_ONCE(qlock->lock)); + /* + * Set the lock bit & clear the waiting bit + */ + if (cmpxchg(&qlock->lock_wait, _QSPINLOCK_WAITING, + _QSPINLOCK_LOCKED) == _QSPIN...

...nlock slowpath. Given that speed at low contention level which is the common case is important to get this patch accepted, I have to do what I can to make it run as far as possible for this 2 contending task case. >> + return 1; >> + } else if (old == _QSPINLOCK_LOCKED) { >> +try_again: >> + /* >> + * Wait until the lock byte is cleared to get the lock >> + */ >> + do { >> + cpu_relax(); >> + } while (ACCESS_ONCE(qlock->lock)); >> + /* >> + * Set the lock bit& clear the waiting bit >> + */ >> + if...

Hi, For EAs data structure in inode/block are little different from them in bucket. These two patches try to make them same for the most part. The first patch set xh_free_start and xh_name_value_len when EAs in inode/block. xh_free_start is useful to keep the minimum offset of the xattr name/value. But xh_name_value_len is not very useful because we don't have "hole" when EAs in

Hi, I have fixed the problems in version 1 patches. These two patches based on the latest main line kernel. Thanks, tiger > For EAs data structure in inode/block are little different from them in > bucket. These two patches try to make them same for the most part. > > The first patch set xh_free_start and xh_name_value_len when EAs in > inode/block. xh_free_start is useful to

When the extended attribute namespace grows to a b-tree, the leaf clusters are organized by means of 'buckets'. Each bucket is 4K in size, regardless of blocksize. Thus, a bucket may be made of more than one block. fs/ocfs2/xattr.c has a nice little abstraction to wrap this, struct ocfs2_xattr_bucket. It contains a list of buffer_heads representing these blocks, and there is even an

...Component: Miscellaneous AssignedTo: openssh-unix-dev at mindrot.org ReportedBy: schwartz+q-bugzilla-mindrot at bio.cse.psu.edu ssh is littered with open coded checks of errno after a read or write. As a matter of good engineering, these should be consolidated into one routine, try_again(errno). The current set of checks (which are quite inconsistent) fail to include all the values of errno that signal that a retry is called for. For example, EHOSTUNREACH, is an advisory rather than an error. Currently ssh will abort a session when TCP/IP is perfectly happy to continue. Because w...

...the lock byte by the lock holder. Utilizing other bits in the low byte for other purpose will complicate the unlock path and slow down the no-contention case. >>>> + /* >>>> + * Someone has steal the lock, so wait again >>>> + */ >>>> + goto try_again; >>> That's just a fail.. steals should not ever be allowed. It's a fair lock >>> after all. >> The code is unfair, but this unfairness help it to run faster than ticket >> spinlock in this particular case. And the regular qspinlock slowpath is >> fair. A...

...->xh; + u32 name_hash = ocfs2_xattr_name_hash(inode, name, strlen(name)); + + if (name_hash != le32_to_cpu(xh->xh_entries[0].xe_name_hash)) + return 0; if (xh->xh_entries[le16_to_cpu(xh->xh_count) - 1].xe_name_hash == xh->xh_entries[0].xe_name_hash) { @@ -4562,7 +4645,9 @@ try_again: * one bucket's worth, so check it here whether we need to * add a new bucket for the insert. */ - ret = ocfs2_check_xattr_bucket_collision(inode, &xs->bucket); + ret = ocfs2_check_xattr_bucket_collision(inode, + &xs->bucket, + xi->name); if (ret)...

...e thing with a single cmpxchg. No extra ops needed. And at that point you don't need a whole byte, you can use a single bit. that removes the whole NR_CPUS dependent logic. > >>+ /* > >>+ * Someone has steal the lock, so wait again > >>+ */ > >>+ goto try_again; > >That's just a fail.. steals should not ever be allowed. It's a fair lock > >after all. > > The code is unfair, but this unfairness help it to run faster than ticket > spinlock in this particular case. And the regular qspinlock slowpath is > fair. A little bit o...

...e thing with a single cmpxchg. No extra ops needed. And at that point you don't need a whole byte, you can use a single bit. that removes the whole NR_CPUS dependent logic. > >>+ /* > >>+ * Someone has steal the lock, so wait again > >>+ */ > >>+ goto try_again; > >That's just a fail.. steals should not ever be allowed. It's a fair lock > >after all. > > The code is unfair, but this unfairness help it to run faster than ticket > spinlock in this particular case. And the regular qspinlock slowpath is > fair. A little bit o...

...,7 @@ static int ehci_wait_for_port(int port); * Return -ENODEV for any general failure * Return -EIO if wait for port fails */ -int dbgp_external_startup(void) +static int _dbgp_external_startup(void) { int devnum; struct usb_debug_descriptor dbgp_desc; @@ -613,6 +613,11 @@ err: goto try_again; return -ENODEV; } + +int dbgp_external_startup(struct usb_hcd *hcd) +{ + return xen_dbgp_external_startup(hcd) ?: _dbgp_external_startup(); +} EXPORT_SYMBOL_GPL(dbgp_external_startup); static int ehci_reset_port(int port) @@ -804,7 +809,7 @@ try_next_port: dbgp_ehci_status("ehci ski...

I am building a application for my university and I would like to make this application authenticate agains a Active Directory server. I would like to know if there is an already built solution for this somewhere online ? Thanks -- Nicolas Kassis -------------------- http://www.nickassis.net http://www.nickassis.net/blog -------------- next part -------------- An HTML attachment was scrubbed...

Another bundle of issues from Coverity triage. The first one is in x86/mm, and looks scarier than it is. The others are all in xen/drivers and AFAICT are pretty minor. Cheers, Tim.

I tried to compile Samba 2.0 beta with cc under SCO 3.2. I seem to be missing the inet.h and netdb.h include files. Does anyone know if they are publicly available? Steve Grose Sgrose@cmps.com Continental Managed Pharmacy Services - www.preferrx.com Voice - 216-459-2025 Ext. 208 Fax - 216-485-8615 Any opinions expressed are my own and not necessarily those of my employers.

...ile html test_package html test_path html test_that html testthat html testthat_results html try_again html use_catch html watch html with_mock html ** building package indices ** testing if installed package can be loaded * DONE (testthat) Making 'packa...