<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Recent changes to bugs</title><link>https://sourceforge.net/p/hbaapi/bugs/</link><description>Recent changes to bugs</description><atom:link href="https://sourceforge.net/p/hbaapi/bugs/feed.rss" rel="self"/><language>en</language><lastBuildDate>Tue, 03 Nov 2009 22:02:48 -0000</lastBuildDate><atom:link href="https://sourceforge.net/p/hbaapi/bugs/feed.rss" rel="self" type="application/rss+xml"/><item><title>makefile correction needed on unix</title><link>https://sourceforge.net/p/hbaapi/bugs/15/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;The libHBAAPI.so has a library dependency on libdl.so because of the "dlopen" function call it makes. However, the makefile.unix has the "-ldl" linker option in hbatest.c not libhbaapi.so. Here is a copy of the relevant lines in the makefile:&lt;/p&gt;
&lt;p&gt;hbatest: hbaapi.h&lt;br /&gt;
    ${CC} hbatest.c -ldl -o hbatest&lt;/p&gt;
&lt;p&gt;libHBAAPI.so: HBAAPILIB.c hbaapi.h vendorhbaapi.h&lt;br /&gt;
    ${VERS}&lt;br /&gt;
    ${CC} HBAAPILIB.c  version.o ${CFLAGS} -DBUILD_DATE=\"'${BUILD_DATE}'\" -G -o libHBAAPI.so&lt;br /&gt;
    @rm version.o&lt;/p&gt;
&lt;p&gt;Thus, the -ldl needs be moved out of hbatest.c logic shown above and into the libHBAAPI build logic. Because this is not currently done, the libHBAAPI.so will work with the HBAAPI client apps this project is building (e.g. hbatest.c) but will not work when the libHBAAPI.so is built with other (3rd party) apps who don't use -ldl. Again, the issue here is that libHBAAPI.so should include the -ldl (link with libdl.so) compiler option because it is the code that make the dlopen (libdl.so) function call. We had to fix our 3rd party client to include -ldl in our client makefile as a workaround so we could use this libHBAAPI shared object library. The linux "ldd" command can be run on the hbaapi executable and shared object binary files to also verify this issue.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">chuck martin</dc:creator><pubDate>Tue, 03 Nov 2009 22:02:48 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net65993c13ad1a3242cfd574c1568e5292d35b6564</guid></item><item><title>stray printf spews to stdout</title><link>https://sourceforge.net/p/hbaapi/bugs/14/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;line 679 in HBAAPILIB.c performs a printf in an error condition. This results in writing to stdout and is undesired for applications / daeamons which use their own logging mechanisms ( files &amp;amp; syslog ).&lt;/p&gt;
&lt;p&gt;Simple fix is to comment out this line.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Peter Mellquist</dc:creator><pubDate>Mon, 29 Jan 2007 23:04:48 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net98336cb5a65aa02720fb90849c845cc251ff54f5</guid></item><item><title>Deadlocks with HbaApiLib.c</title><link>https://sourceforge.net/p/hbaapi/bugs/13/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;We ran into a deadlock situation with multiple threads &lt;br /&gt;
while using particularly port state event callbacks, but &lt;br /&gt;
my guess is that this can happen with just about any of &lt;br /&gt;
the registration / callback handlers.  The one we were &lt;br /&gt;
using was adapterportstateevents_callback, which had &lt;br /&gt;
another bug previously reported.&lt;/p&gt;
&lt;p&gt;Now, what happened is that the application's main &lt;br /&gt;
thread was registering for these callbacks in a multiple &lt;br /&gt;
adapter installation on a HyperThreading capable &lt;br /&gt;
processor.  As soon as the first adapter's callback got &lt;br /&gt;
registered, a callback occurred in another threadd on &lt;br /&gt;
this adapter and IT locked the APSE mutex.  Our &lt;br /&gt;
callback function in turn called another HBA API function &lt;br /&gt;
that attempted to lock the LL mutex.  But the main &lt;br /&gt;
thread was in the process of registering callbacks and IT &lt;br /&gt;
had the LL mutex locked and was trying to get the APSE &lt;br /&gt;
mutex that the other thread had locked.&lt;/p&gt;
&lt;p&gt;OUR solution was to first lock the LL mutex in any of the &lt;br /&gt;
callback functions that lock the other mutexes.&lt;/p&gt;
&lt;p&gt;I don't recall seeing anything in the documentation that &lt;br /&gt;
said calling back into the HBA library was illegal from a &lt;br /&gt;
callback function.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jerry Masker</dc:creator><pubDate>Mon, 18 Oct 2004 22:56:08 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net080af6348e538e6d4176536dfccaf897c9dc9d30</guid></item><item><title>Lost mutex after bad handle</title><link>https://sourceforge.net/p/hbaapi/bugs/12/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Function HBA_CheckLibrary, HbaApiLib.c, line 339.&lt;br /&gt;
Admittedly this should only happen if the user passes in &lt;br /&gt;
a bad handle referencing a library not found, but the line &lt;br /&gt;
in question should be a RELEASE_MUTEX_RETURN &lt;br /&gt;
instead of just return.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jerry Masker</dc:creator><pubDate>Mon, 18 Oct 2004 22:45:21 -0000</pubDate><guid isPermaLink="false">https://sourceforge.netc0aefa3652e0c4d16fc9b9c22a3eaed0707dc69c</guid></item><item><title>Locked mutex after callback</title><link>https://sourceforge.net/p/hbaapi/bugs/11/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Function adapterportstateevents_callback, HbaApiLib.c, &lt;br /&gt;
line 1674.  This return needs to be a break, and you &lt;br /&gt;
need a RELEASE_MUTEX on _hbaapi_APSE_mutex before &lt;br /&gt;
the function exits.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Anonymous</dc:creator><pubDate>Mon, 18 Oct 2004 22:24:09 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net245e631795ed82f0ab46cd55083593bbd5903a3c</guid></item><item><title>No matching fclose in HBA_LoadLibrary</title><link>https://sourceforge.net/p/hbaapi/bugs/10/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;There is an fopen of hbaConfFilePath in HBA_LoadLibrary &lt;br /&gt;
for which there is no matching fclose.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Anonymous</dc:creator><pubDate>Wed, 17 Mar 2004 22:18:28 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net5ff36e717f09db2886a56a4122d6089a936a0e82</guid></item><item><title>License agreement URL is bad.</title><link>https://sourceforge.net/p/hbaapi/bugs/9/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;The URL indicated in the HBA API license agreement&lt;br /&gt;
file &lt;br /&gt;
hbaapi/hbaapi/license.htm seems to no longer be correct. It &lt;br /&gt;
is:&lt;/p&gt;
&lt;p&gt;&lt;a href="http://www.snia.org/English/Resources/Code/OpenSource.html" rel="nofollow"&gt;http://www.snia.org/English/Resources/Code/OpenSource.html&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;I &lt;br /&gt;
expect that a SNIA site redesign misplaced the document. I &lt;br /&gt;
searched the site for it, but did not find it &lt;/p&gt;
&lt;p&gt;anywhere.&lt;/p&gt;
&lt;p&gt;~ Jeff Byers ~&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jeff Byers</dc:creator><pubDate>Tue, 30 Jul 2002 19:33:50 -0000</pubDate><guid isPermaLink="false">https://sourceforge.netb153a94dd5bfd3a9040e5a1b34e59c26df7f60eb</guid></item><item><title>Deadlock caused by RegisterForTargetEv</title><link>https://sourceforge.net/p/hbaapi/bugs/8/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;In &lt;br /&gt;
HBA_STATUS&lt;br /&gt;
HBA_RegisterForTargetEvents (&lt;br /&gt;
void        (*callback) (&lt;br /&gt;
    void        *data,&lt;br /&gt;
    HBA_WWN  &lt;br /&gt;
    hbaPortWWN,&lt;br /&gt;
    HBA_WWN  &lt;br /&gt;
    discoveredPortWWN,&lt;br /&gt;
    HBA_UINT32 eventType&lt;br /&gt;
    ),&lt;/p&gt;
&lt;p&gt;The  _hbaapi_LL_mutex is grabbed by calling &lt;br /&gt;
CheckLibrary().  This mutex is never released if the call &lt;br /&gt;
is successful, causing a deadlock when any of the event &lt;br /&gt;
callbacks are fired.  All of the other RegisterForXXX &lt;br /&gt;
functions end with the line:&lt;/p&gt;
&lt;p&gt;RELEASE_MUTEX_RETURN(&amp;amp;amp;_hbaapi_LL_mutex, &lt;br /&gt;
HBA_STATUS_OK);&lt;/p&gt;
&lt;p&gt;RegisterForTargetEvents simply returns the status and &lt;br /&gt;
doesn't release the mutex (~ line 1850):&lt;/p&gt;
&lt;p&gt;GRAB_MUTEX(&amp;amp;amp;_hbaapi_TE_mutex);&lt;br /&gt;
acbp-&amp;amp;gt;next = _hbaapi_targetevents_callback_list;&lt;br /&gt;
_hbaapi_targetevents_callback_list = acbp;&lt;br /&gt;
RELEASE_MUTEX(&amp;amp;amp;_hbaapi_TE_mutex);&lt;/p&gt;
&lt;p&gt;return HBA_STATUS_OK;&lt;br /&gt;
}&lt;/p&gt;
&lt;p&gt;adding the RELEASE_MUTEX... line removes the &lt;br /&gt;
deadlock.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Todd Flanagan</dc:creator><pubDate>Wed, 12 Jun 2002 18:45:18 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net783df3532168bc3bbca5cc99a674bf85518bdc41</guid></item><item><title>DEBUG macro has syntax errors</title><link>https://sourceforge.net/p/hbaapi/bugs/7/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;The DEBUG macro will not compile when USELOGFILE &lt;br /&gt;
is defined.  The wwn2Str is also being used incorrectly &lt;br /&gt;
in the macro itself.  Finally, the temp filename is &lt;br /&gt;
incorrect for the NT platform.  See attached file for &lt;br /&gt;
enough basic fixes to get the library to compile and &lt;br /&gt;
produce a log file.  It needs some refactoring, but it gets &lt;br /&gt;
the job done for now.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Todd Flanagan</dc:creator><pubDate>Fri, 07 Jun 2002 14:49:06 -0000</pubDate><guid isPermaLink="false">https://sourceforge.net5c412053b293ed62f308fe0bfe92b69b3e1ba22b</guid></item><item><title>sample library doesn't compile</title><link>https://sourceforge.net/p/hbaapi/bugs/6/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Not even close.  Obviously never event tried in v2.1.&lt;br /&gt;
Attached are enough fixes to get past compilation.&lt;br /&gt;
Below is the header, attached is the .c&lt;/p&gt;
&lt;p&gt;/*&lt;br /&gt;
* Description&lt;br /&gt;
*         hbasample.h - Header file for sample HBA API &lt;br /&gt;
compliant vendor library&lt;br /&gt;
*&lt;br /&gt;
* License:&lt;br /&gt;
*  The contents of this file are subject to the &lt;br /&gt;
SNIA Public License&lt;br /&gt;
*  Version 1.0 (the &amp;amp;quot;License&amp;amp;quot;); you may not use &lt;br /&gt;
this file except in&lt;br /&gt;
*  compliance with the License. You may &lt;br /&gt;
obtain a copy of the License at&lt;br /&gt;
*&lt;br /&gt;
*&lt;br /&gt;
    /http://www.snia.org/English/Resources/Code/&lt;br /&gt;
OpenSource.html&lt;br /&gt;
*&lt;br /&gt;
*  Software distributed under the License is &lt;br /&gt;
distributed on an &amp;amp;quot;AS IS&amp;amp;quot;&lt;br /&gt;
*  basis, WITHOUT WARRANTY OF ANY &lt;br /&gt;
KIND, either express or implied. See&lt;br /&gt;
*  the License for the specific language &lt;br /&gt;
governing rights and limitations&lt;br /&gt;
*  under the License.&lt;br /&gt;
*&lt;br /&gt;
* The Original Code is  Header file for&lt;br /&gt;
*   SNIA HBA API Sample Vendor Specific &lt;br /&gt;
Library&lt;br /&gt;
*&lt;br /&gt;
* The Initial Developer of the Original Code is:&lt;br /&gt;
*  Benjamin F. Kuo, Troika Networks, Inc. &lt;br /&gt;
(benk@troikanetworks.com)&lt;br /&gt;
*&lt;br /&gt;
* Contributor(s):&lt;br /&gt;
*&lt;/p&gt;
&lt;p&gt;******************************************************************&lt;br /&gt;
********* &lt;br /&gt;
*/&lt;br /&gt;
#ifdef WIN32&lt;br /&gt;
#ifdef HBASAMPLE_EXPORTS&lt;br /&gt;
#define HBASAMPLE_API __declspec(dllexport)&lt;br /&gt;
#else&lt;br /&gt;
#define HBASAMPLE_API __declspec(dllimport)&lt;br /&gt;
#endif&lt;br /&gt;
#else&lt;br /&gt;
#define HBASAMPLE_API&lt;br /&gt;
#endif&lt;/p&gt;
&lt;p&gt;HBASAMPLE_API HBA_STATUS HBA_RegisterLibrary( &lt;br /&gt;
PHBA_ENTRYPOINTS entrypoints );&lt;/p&gt;
&lt;p&gt;HBASAMPLE_API HBA_STATUS&lt;br /&gt;
HBA_RegisterLibraryV2(HBA_ENTRYPOINTSV2 &lt;br /&gt;
*entrypoints);&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_LoadLibrary();&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_FreeLibrary();&lt;/p&gt;
&lt;p&gt;HBA_UINT32 Sample_GetVersion();&lt;/p&gt;
&lt;p&gt;HBA_UINT32 Sample_GetNumberOfAdapters();&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_GetAdapterName(HBA_UINT32 &lt;br /&gt;
adapterindex, char *adaptername);&lt;/p&gt;
&lt;p&gt;HBA_HANDLE&lt;br /&gt;
Sample_OpenAdapter(char* adaptername );&lt;/p&gt;
&lt;p&gt;void Sample_CloseAdapter(HBA_HANDLE handle);&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_GetAdapterAttributes(&lt;br /&gt;
    HBA_HANDLE handle, &lt;br /&gt;
    PHBA_ADAPTERATTRIBUTES hbaattributes&lt;/p&gt;
&lt;div class="codehilite"&gt;&lt;pre&gt;&lt;span class="o"&gt;\&lt;/span&gt;&lt;span class="p"&gt;);&lt;/span&gt;
&lt;/pre&gt;&lt;/div&gt;
&lt;p&gt;HBA_STATUS Sample_GetAdapterPortAttributes(&lt;br /&gt;
    HBA_HANDLE handle, &lt;br /&gt;
    HBA_UINT32 portindex,&lt;br /&gt;
    HBA_PORTATTRIBUTES *portattributes&lt;br /&gt;
    );&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_GetPortStatistics(&lt;br /&gt;
    HBA_HANDLE       &lt;br /&gt;
    handle,&lt;br /&gt;
    HBA_UINT32       &lt;br /&gt;
    portindex,&lt;br /&gt;
    HBA_PORTSTATISTICS   &lt;br /&gt;
    *portstatistics&lt;br /&gt;
    );&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_GetDiscoveredPortAttributes(&lt;br /&gt;
    HBA_HANDLE handle, &lt;br /&gt;
    HBA_UINT32 portindex,&lt;br /&gt;
    HBA_UINT32 discoveredportindex,&lt;br /&gt;
    HBA_PORTATTRIBUTES *portattributes&lt;br /&gt;
    );&lt;br /&gt;
HBA_STATUS Sample_GetPortAttributesByWWN(&lt;br /&gt;
    HBA_HANDLE handle,&lt;br /&gt;
    HBA_WWN PortWWN,&lt;br /&gt;
    HBA_PORTATTRIBUTES *portattributes&lt;br /&gt;
    );&lt;/p&gt;
&lt;p&gt;void Sample_RefreshInformation(HBA_HANDLE handle);&lt;/p&gt;
&lt;p&gt;typedef void (*callback_t)(void*, HBA_WWN, &lt;br /&gt;
HBA_UINT32);&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_RegisterForAdapterEvents(&lt;br /&gt;
callback_t callback,&lt;br /&gt;
void *userData,&lt;br /&gt;
HBA_HANDLE handle,&lt;br /&gt;
HBA_CALLBACKHANDLE *callbackHandle);&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_HBAOpenAdapterByWWN&lt;br /&gt;
(HBA_HANDLE *phandle, HBA_WWN wwn);&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_HBA_GetFcpTargetMapping(&lt;br /&gt;
HBA_HANDLE     handle,&lt;br /&gt;
HBA_FCPTARGETMAPPING&lt;br /&gt;
            *pmapping&lt;br /&gt;
);&lt;/p&gt;
&lt;p&gt;HBA_STATUS Sample_SendCTPassThruV2&lt;br /&gt;
(HBA_HANDLE, HBA_WWN, void *, HBA_UINT32, void &lt;br /&gt;
*, HBA_UINT32 *);&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Todd Flanagan</dc:creator><pubDate>Tue, 04 Jun 2002 20:21:15 -0000</pubDate><guid isPermaLink="false">https://sourceforge.netd9fec5ecc05eeb48d1210039e80c79e0e458ea52</guid></item></channel></rss>