| Message ID | 1310590312-21669-2-git-send-email-ben@bwidawsk.net (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, 13 Jul 2011 13:51:43 -0700, Ben Widawsky <ben@bwidawsk.net> wrote: > > Signed-off-by: Ben Widawsky <ben@bwidawsk.net> > --- > intel/Makefile.am | 3 ++- > intel/intel_debug.h | 44 ++++++++++++++++++++++++++++++++++++++++++++ > 2 files changed, 46 insertions(+), 1 deletions(-) > > diff --git a/intel/Makefile.am b/intel/Makefile.am > index b6a9014..7a44aaf 100644 > --- a/intel/Makefile.am > +++ b/intel/Makefile.am > @@ -45,6 +45,7 @@ libdrm_intel_la_SOURCES = \ > mm.h > > libdrm_intelincludedir = ${includedir}/libdrm > -libdrm_intelinclude_HEADERS = intel_bufmgr.h > +libdrm_intelinclude_HEADERS = intel_bufmgr.h \ > + intel_debug.h > > pkgconfig_DATA = libdrm_intel.pc > diff --git a/intel/intel_debug.h b/intel/intel_debug.h > new file mode 100644 > index 0000000..42ea1d8 > --- /dev/null > +++ b/intel/intel_debug.h > @@ -0,0 +1,44 @@ > +/* > + * Copyright © 2011 Intel Corporation > + * > + * Permission is hereby granted, free of charge, to any person obtaining a > + * copy of this software and associated documentation files (the "Software"), > + * to deal in the Software without restriction, including without limitation > + * the rights to use, copy, modify, merge, publish, distribute, sublicense, > + * and/or sell copies of the Software, and to permit persons to whom the > + * Software is furnished to do so, subject to the following conditions: > + * > + * The above copyright notice and this permission notice (including the next > + * paragraph) shall be included in all copies or substantial portions of the > + * Software. > + * > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL > + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER > + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING > + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS > + * IN THE SOFTWARE. > + * > + * Authors: > + * Ben Widawsky <ben@bwidawsk.net> > + * > + */ > + > +#ifndef INTEL_DEBUG_H > +#define INTEL_DEBUG_H > + > +#include <stdint.h> > + > +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug" > +#define DEBUG_HANDSHAKE_VERSION 0x2 > +#define DEBUG_HANDSHAKE_ACK "okay" > + > +/* First byte must always be the 1 byte version */ > +struct debug_handshake { > + uint8_t version; > + int flink_handle; > + uint32_t per_thread_scratch; > +} __attribute__((packed)); Let's use an intel_ namespace to avoid any potential confusion. Do you want to reserve the first uint32_t for your protocol header? Then it at least looks like the hole will be useful in future... I'm not overly thrilled by intel_debug.h, it's so generic but I can't think of anything better. -Chris
On Wed, Jul 13, 2011 at 13:51:43 -0700, Ben Widawsky wrote:
> +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug"
Not sure what this is used for, but does it really need to be in /tmp?
Cheers,
Julien
On Tue, Jul 19, 2011 at 11:06:17PM +0200, Julien Cristau wrote: > On Wed, Jul 13, 2011 at 13:51:43 -0700, Ben Widawsky wrote: > > > +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug" > > Not sure what this is used for, but does it really need to be in /tmp? It is the shared socket between a debug client (ie. Mesa) and the debugger. I don't care where it goes, do you have any recommendations? > > Cheers, > Julien Thanks. Ben
On Thu, Jul 21, 2011 at 13:54:41 +0000, Ben Widawsky wrote: > On Tue, Jul 19, 2011 at 11:06:17PM +0200, Julien Cristau wrote: > > On Wed, Jul 13, 2011 at 13:51:43 -0700, Ben Widawsky wrote: > > > > > +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug" > > > > Not sure what this is used for, but does it really need to be in /tmp? > > It is the shared socket between a debug client (ie. Mesa) and the > debugger. I don't care where it goes, do you have any recommendations? > Somewhere under $HOME is probably better than a predictable file name in /tmp. Cheers, Julien
On Thu, Jul 21, 2011 at 11:22:12PM +0200, Julien Cristau wrote: > On Thu, Jul 21, 2011 at 13:54:41 +0000, Ben Widawsky wrote: > > > On Tue, Jul 19, 2011 at 11:06:17PM +0200, Julien Cristau wrote: > > > On Wed, Jul 13, 2011 at 13:51:43 -0700, Ben Widawsky wrote: > > > > > > > +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug" > > > > > > Not sure what this is used for, but does it really need to be in /tmp? > > > > It is the shared socket between a debug client (ie. Mesa) and the > > debugger. I don't care where it goes, do you have any recommendations? > > > Somewhere under $HOME is probably better than a predictable file name in > /tmp. $HOME is not great because it requires a little hackery to accomplish. The debugger will be running as root, and so if we put it in root's $HOME, mesa may not be able to reach it. > > Cheers, > Julien Ben
On Wed, Jul 27, 2011 at 15:08:08 +0000, Ben Widawsky wrote: > On Thu, Jul 21, 2011 at 11:22:12PM +0200, Julien Cristau wrote: > > On Thu, Jul 21, 2011 at 13:54:41 +0000, Ben Widawsky wrote: > > > > > On Tue, Jul 19, 2011 at 11:06:17PM +0200, Julien Cristau wrote: > > > > On Wed, Jul 13, 2011 at 13:51:43 -0700, Ben Widawsky wrote: > > > > > > > > > +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug" > > > > > > > > Not sure what this is used for, but does it really need to be in /tmp? > > > > > > It is the shared socket between a debug client (ie. Mesa) and the > > > debugger. I don't care where it goes, do you have any recommendations? > > > > > Somewhere under $HOME is probably better than a predictable file name in > > /tmp. > > $HOME is not great because it requires a little hackery to accomplish. > The debugger will be running as root, and so if we put it in root's > $HOME, mesa may not be able to reach it. > Ah, I didn't realise this would run as root. Would /run (or /var/run) be ok then? Alternately, use an abstract domain socket to not have to care about the underlying filesystem. Cheers, Julien
> Ah, I didn't realise this would run as root. Would /run (or /var/run) > be ok then? Alternately, use an abstract domain socket to not have to > care about the underlying filesystem. Abstract sockets are a bit of a Linux specific feature. You also don't want sockets in /home as it may well be NFS mounted on some configurations and AF_UNIX sockets don't mix with NFS. /var/run would probably be a lot more sensible for a system debug interface I agree. Alan
On Wed, Jul 27, 2011 at 05:16:31PM +0200, Julien Cristau wrote: > On Wed, Jul 27, 2011 at 15:08:08 +0000, Ben Widawsky wrote: > > > On Thu, Jul 21, 2011 at 11:22:12PM +0200, Julien Cristau wrote: > > > On Thu, Jul 21, 2011 at 13:54:41 +0000, Ben Widawsky wrote: > > > > > > > On Tue, Jul 19, 2011 at 11:06:17PM +0200, Julien Cristau wrote: > > > > > On Wed, Jul 13, 2011 at 13:51:43 -0700, Ben Widawsky wrote: > > > > > > > > > > > +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug" > > > > > > > > > > Not sure what this is used for, but does it really need to be in /tmp? > > > > > > > > It is the shared socket between a debug client (ie. Mesa) and the > > > > debugger. I don't care where it goes, do you have any recommendations? > > > > > > > Somewhere under $HOME is probably better than a predictable file name in > > > /tmp. > > > > $HOME is not great because it requires a little hackery to accomplish. > > The debugger will be running as root, and so if we put it in root's > > $HOME, mesa may not be able to reach it. > > > Ah, I didn't realise this would run as root. Would /run (or /var/run) > be ok then? Alternately, use an abstract domain socket to not have to > care about the underlying filesystem. I wasn't even aware of abstract domain sockets. Seems like fun, let me try that one first. Ben
diff --git a/intel/Makefile.am b/intel/Makefile.am index b6a9014..7a44aaf 100644 --- a/intel/Makefile.am +++ b/intel/Makefile.am @@ -45,6 +45,7 @@ libdrm_intel_la_SOURCES = \ mm.h libdrm_intelincludedir = ${includedir}/libdrm -libdrm_intelinclude_HEADERS = intel_bufmgr.h +libdrm_intelinclude_HEADERS = intel_bufmgr.h \ + intel_debug.h pkgconfig_DATA = libdrm_intel.pc diff --git a/intel/intel_debug.h b/intel/intel_debug.h new file mode 100644 index 0000000..42ea1d8 --- /dev/null +++ b/intel/intel_debug.h @@ -0,0 +1,44 @@ +/* + * Copyright © 2011 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + * Ben Widawsky <ben@bwidawsk.net> + * + */ + +#ifndef INTEL_DEBUG_H +#define INTEL_DEBUG_H + +#include <stdint.h> + +#define SHADER_DEBUG_SOCKET "/tmp/gen_debug" +#define DEBUG_HANDSHAKE_VERSION 0x2 +#define DEBUG_HANDSHAKE_ACK "okay" + +/* First byte must always be the 1 byte version */ +struct debug_handshake { + uint8_t version; + int flink_handle; + uint32_t per_thread_scratch; +} __attribute__((packed)); + +#endif
Signed-off-by: Ben Widawsky <ben@bwidawsk.net> --- intel/Makefile.am | 3 ++- intel/intel_debug.h | 44 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 46 insertions(+), 1 deletions(-)