libxl: prepare environment for domcreate_stream_done
PATCH v2 1/1 libxl: prepare environment for domcreate_stream_done
The function domcreate_bootloader_done may branch early to domcreate_stream_done, in case some error occoured. Here srs->dcs will be NULL, which leads to a crash.
It is unclear what the purpose of that backpointer is. Perhaps it can be removed, and domcreate_stream_done could use CONTAINER_OF.
---
v2:
unconditional assignment
tools/libxl/libxl_create.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
tools/libxl/libxl_create.c +3 -1
@@ -1093,6 +1093,9 @@ static void domcreate_bootloader_done(libxl__egc *egc,return ; } +/* Prepare environment for domcreate_stream_done */ + dcs->srs.dcs = dcs; +/* Restore */ callbacks->restore_results = libxl__srm_callout_callback_restore_results;
@@ -1116,7 +1119,6 @@ static void domcreate_bootloader_done(libxl__egc *egc, goto out; dcs->srs.ao = ao; - dcs->srs.dcs = dcs; dcs->srs.fd = restore_fd; dcs->srs.legacy = (dcs->restore_params.stream_version ==1 ); dcs->srs.back_channel = false;
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Quoted 8 lines · Olaf Hering
The function
domcreate_bootloader_donemay branch early todomcreate_stream_done, in case some error occoured. Heresrs->dcswill be NULL, which leads to a crash.It is unclear what the purpose of that backpointer is. Perhaps it can be removed, and
domcreate_stream_donecould useCONTAINER_OF.Signed-off-by Olaf Hering <[email protected]>
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Olaf Hering writes ("[PATCH v2] libxl: prepare environment for domcreate_stream_done"):
Quoted 3 lines
The function
domcreate_bootloader_donemay branch early todomcreate_stream_done, in case some error occoured. Heresrs->dcswill be NULL, which leads to a crash.
Thanks. I think this is OK as far as it goes. But:
tools/libxl/libxl_create.c static void domcreate_bootloader_done(libxl__egc *egc, · line 1097
Quoted 2 lines
tools/libxl/libxl_create.c +2
-0+/* Prepare environment for domcreate_stream_done */ + dcs->srs.dcs = dcs;
The need for this comment is telling us something about the weird code structure here. We initialise all of dcs->srs much later, so even with this change we call domcreate_stream_done with a mostly-uninitialised srs.
In fact this is OK because domcreate_stream_done only reads srs->dcs and then does everything with the obtained dcs. But there is nothing there to indicate that srs might be mostly uninitialised. Maybe we could add a comment there, something like:
/* NB perhaps only srs->dcs is valid; eg in the case of an
* early branch to domcreate_bootloader_done's `out' block */
?
Ian.
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Am Fri, 8 Mar 2019 14:08:10 +0000
schrieb Ian Jackson <[email protected]>:
Quoted 7 lines
In fact this is OK because
domcreate_stream_doneonly readssrs->dcsand then does everything with the obtained dcs. But there is nothing there to indicate that srs might be mostly uninitialised. Maybe we could add a comment there, something like:/* NB perhaps only
srs->dcsis valid; eg in the case of an
* early branch to domcreate_bootloader_done's `out' block */
I'm find with that. Can this comment be adjusted at commit time?
Olaf
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Quoted 8 lines · Olaf Hering
Am Fri, 8 Mar 2019 14:08:10 +0000
schrieb Ian Jackson <[email protected]>:Quoted 5 lines
In fact this is OK because
domcreate_stream_doneonly readssrs->dcsand then does everything with the obtained dcs. But there is nothing there to indicate that srs might be mostly uninitialised. Maybe we could add a comment there, something like:
Quoted 4 lines
Quoted 2 lines
/* NB perhaps only
srs->dcsis valid; eg in the case of an
* early branch to domcreate_bootloader_done's `out' block */I'm find with that. Can this comment be adjusted at commit time?
Sure.
Wei.
Quoted 2 lines
Olaf
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Olaf Hering writes ("Re: [PATCH v2] libxl: prepare environment for domcreate_stream_done"):
Quoted 8 lines
Am Fri, 8 Mar 2019 14:08:10 +0000
schrieb Ian Jackson <[email protected]>:Quoted 5 lines
In fact this is OK because
domcreate_stream_doneonly readssrs->dcsand then does everything with the obtained dcs. But there is nothing there to indicate that srs might be mostly uninitialised. Maybe we could add a comment there, something like:
Quoted 4 lines
Quoted 2 lines
/* NB perhaps only
srs->dcsis valid; eg in the case of an
* early branch to domcreate_bootloader_done's `out' block */I'm find with that. Can this comment be adjusted at commit time?
Sure, although if it's me that commits it I'd do it as a followup patch.
We need a release ack I think for 4.12. Subject adjusted, CCing Juergen. Juergen, will bounce the original patch to you too.
Ian.
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Quoted 8 lines · Olaf Hering
The function
domcreate_bootloader_donemay branch early todomcreate_stream_done, in case some error occoured. Heresrs->dcswill be NULL, which leads to a crash.It is unclear what the purpose of that backpointer is. Perhaps it can be removed, and
domcreate_stream_donecould useCONTAINER_OF.Signed-off-by Olaf Hering <[email protected]>
As we are very close to a 4.12 release and this is no recent regression I'd like to defer this patch to 4.13.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel