The old StoreData RPC is only supposed to be valid for files
smaller than 2GB. When StoreData64 was added, StoreData and StoreData64
were implemented as calls to common_StoreData64. This removed the bounds
checking on the old StoreData RPC making it possible for operations
beyond two 2GB to be requested even if the file server cannot support
them. This patch adds a validity check to ensure that the requested
file operations remain below 2GB. Failures return E2BIG.
A number of crash reports have been filed with Microsoft in afspthread.dll.
The crashes are occuring as the processes are terminating. In order to
attempt to address this problem this patch adds support for a DllMain()
entry point that is used to cleanup Thread Local Storage and the various
RX queues. One theory is that processes are loading and unloading DLLs
that are linked to afspthread.dll leaving the pthread data in a very
inconsistent state after each unload. By cleaning up the static data
during the unload, if afspthread is loading again it will be forced to
initialize the data once again.
This patch adds code to perform return code error checking for all calls
to the ReadXyz() routines. If the rx connection was lost or if the dump
format being parsed was incorrect there was an opportunity for data to
the processed after an error had already been detected. In the best case
this would have simply resulted in additional work being performed when
it was not necessary but there was also the potential for modifying return
values which would later be used as input to later operations. Better
to simply catch the error when it initially occurs and be done with it.
The ReadString() function had the potential to write into one byte prior
to the start of the input array.
ProcessIndex() would not properly close Fdhandle nor the StreamHandle
if an error was detected.
RestoreVolume() had a problem with the processing of incremental dumps.
Originally AFS supported a concept of incrementals dumps which when
restored would simply add their contents to the existing volume. One
of the distinctions between the incremental and full dumps is that full
dump processing is responsible for removing any vnode entries from the
volume that are not present in the dump file headers. When the incremental
dump support was removed, the RestoreVolume() function was broken. If
a dump file contains multiple Dump Headers, then only the vnode entries
in the first Dump Header would be marked as being restored. Any vnode
entries restored from subsequent Dump Headers would be deleted when the
RestoreVolume() function reached the end. By removing the assignment
of 'tdelo = -1' within the loop that processes the Dump Headers, this
bug is resolved. Although RestoreVolume is only called with the
'incremental' parameter == 0 at the current time, the check for non-zero
values is restored and the "remove unrestored vnodes" functionality is
disabled when a non-zero value is provided.
The ReadVnodes() function did not perform error checking.
The 'active_vcp' variable is no longer being used for any useful
purpose and is the center of a race condition that could lead to
an undercount of references to 'vcp' objects. Remove it.
as pointed out by cg2v@andrew.cmu.edu the addition of performance
collection data for the GetCapabilities RPC broke backward compatibility
with previous builds because it altered the size of the data structure
sent on the wire.
Instead we count the number of GetCapabilties calls but do not gather
performance timing data as this can be done using spare fields.
An undercount has been detected of the smb_vc_t objects stored
in the smb_allVCsp list. Unfortunately, we have yet to be able
to find the cause of the undercount so this patch adds logic to
protect against the side effects until such time as the cause
can be identified.
h_Alloc_r was not specifying the port for the newly allocated
host when adding the host to the hash table. Use hashInsert_r
instead so mistakes like this won't happen in the future.
clock_Init() was broken because it would initialize rxi_clock0
with QueryPerformanceCounter() and then call clock_UpdateTime().
So instead of getting an ever increasing value you would get a
fixed value determined by how long it takes to make a function call
between to QueryPerformanceCounter() calls.
If the host obtained during the CallPremable and CallPostamble differ,
then there is the possibility of a hold leak. This patch is a quick
attempt to deal with the potential problem. If the problem occurs
we log it in order to know for sure if this is in fact a problem.
If it is, then at a later date we should perform a more extensive
re-write in which the CallPreamble is used to obtain the host and the
client and pass them both back to the caller which would not be
required to obtain them again at a later date.
In h_Enumerate_r it is possible that the host may be released by the
called procedure. Therefore, we must preserve the pointer to host->next
before the call and hold the host it points to in order to ensure that
it does not disappear out from under us. This ensures that although
the H_LOCK can be dropped during the procedure and the host list can
be rearranged we won't miss more than a handful of hosts during the
enumeration.
when h_GetHost_r replaces 'host' with 'oldHost' we need to release the
hold on 'host' and maintain the hold on 'oldHost' because if a hold
was placed on 'host' in CallPreamble() it will no longer be released
in CallPostamble() since 'oldHost' will now be bound to the client
Implement proper synopsis wrapping for HTML generation.
This was done in three pieces. First, add HTML-specific tags to the POD to
mark the synopsis for HTML purposes so that we can apply style information
to it. Second, update the style sheet to indent all lines except for the
first in the synopsis section. Third, add the appropriate S<> tags around
option and argument pairs so that we don't wrap between the option and its
argument.
Unfortunately, due to the <I<foo>> style that looks nicer for other reasons,
we have to use the very verbose S<<< >>>. Oh well.
SRXAFS_CallBackRxConnAddr calls h_FindClient_r which returns a
client with a refCount and the client->host held. The call
to CallPostamble will release the host hold but not the client
refcount.
Make the mentions of subcommands in the fs command introduction links to
the relevant pages, and add to README a to-do note to do this for the rest
of the introductory pages.
Fix links to man pages that contain underscores by working around a bug
in Pod::Simple.
Do not allow HTBlocks to allocate more blocks than will fit in
hosttablesptrs. Doing so corrupts memory. Panic instead until
we can get this fixed right.
avoid implicit references by including the prototypes from the header except on linux
====================
This delta was composed from multiple commits as part of the CVS->Git migration.
The checkin message with each commit was inconsistent.
The following are the additional commit messages.
====================
FIXES 24765
avoid implicit references by including the prototypes from the header except on
linux
projects / packages / o / openafs.git / log