Skip to contentSkip to conversation

/proc/dri/0/vma broken on nouveau.

3.10rc6 Dave Jones 4 participants 8 messages Jun 18, 2013 – Aug 28, 2013

Subsystems (2)
lore
Share & tools
lore

Dave Jones

author
#0 lore raw

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded. (Note, no X running on that box)

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

------------[ cut here ]------------
kernel BUG at arch/x86/mm/physaddr.c:79!
invalid opcode: 0000 [#1] PREEMPT SMP 
Modules linked in: tun cmtp kernelcapi rfcomm l2tp_ppp l2tp_netlink l2tp_core ipt_ULOG bnep can_bcm scsi_transport_iscsi nfnetlink rds atm bluetooth can_raw appletalk netrom pppoe af_key ipx rose p8023 pppox p8022 psnap ppp_generic ax25 llc slhc af_802154 nfc can rfkill irda crc_ccitt phonet nfsv3 nfs_acl nfs lockd sunrpc fscache nouveau video mxm_wmi wmi i2c_algo_bit ttm drm_kms_helper drm i5000_edac edac_core tg3 iTCO_wdt iTCO_vendor_support lpc_ich ppdev pcspkr microcode mfd_core i2c_i801 i2c_core dcdbas serio_raw ptp i5k_amb pps_core shpchp parport_pc floppy parport xfs libcrc32c raid0
CPU: 3 PID: 26586 Comm: trinity-child3 Tainted: G        W    3.10.0-rc6+ #1 
Hardware name: Dell Inc.                 Precision WorkStation 490    /0DT031, BIOS A08 04/25/2008
task: f12041a0 ti: eace4000 task.ti: eace4000
EIP: 0060:[<c102e2a6>] EFLAGS: 00010206 CPU: 3
EIP is at __phys_addr+0x72/0x76
EAX: 37bfe000 EBX: 37bfe000 ECX: 0000000c EDX: 37bfe000
ESI: eacc7870 EDI: eacc7870 EBP: eace5e98 ESP: eace5e94
 DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
CR0: 80050033 CR2: 0a24116c CR3: 2cc22000 CR4: 000007f0
DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000
DR6: ffff0ff0 DR7: 00000400
Stack:
 ee11ce30 eace5efc f868c42f c106a0ff 80070007 e6266700 00000282 eace5efc
 c15a6ea6 000000d0 f2403680 f2403680 eace5efc c115054d f2d40e50 ee11ce68
 f12041a0 eacc7870 000010b0 eace4000 00001000 c11841cf 000000d0 eacc7870
Call Trace:
 [<f868c42f>] drm_vma_info+0x39/0x26e [drm]
 [<c106a0ff>] ? __might_sleep+0x19/0x1ac
 [<c15a6ea6>] ? mutex_lock_nested+0x286/0x39c
 [<c115054d>] ? kmem_cache_alloc_trace+0x282/0x2cc
 [<c11841cf>] ? seq_read+0x24d/0x438
 [<c1184068>] seq_read+0xe6/0x438
 [<c1183f82>] ? traverse.isra.6+0x256/0x256
 [<c11b9ff6>] proc_reg_read+0x48/0x71
 [<c11b9fae>] ? proc_reg_write+0x71/0x71
 [<c11644c8>] vfs_read+0x80/0x148
 [<c1164b5f>] SyS_read+0x49/0x7d
 [<c15b19c1>] sysenter_do_call+0x12/0x32
Code: 27 11 c2 8d 91 00 00 80 00 39 d0 72 cb 8b 0d 2c 27 81 c1 8d 91 00 00 c1 ff 81 e2 00 00 e0 ff 81 ea 00 20 00 00 39 d0 73 af 0f 0b <0f> 0b 0f 0b 55 89 e5 53 66 66 66 66 90 89 c2 31 c0 81 fa ff ff
#1 lore raw
Quoted 7 lines

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded.
(Note, no X running on that box)

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

How about this, lets just rip it all out.

Dave.

From 54f9605737437272f440bbc6cc4adf805334884b Mon Sep 17 00:00:00 2001
From: Dave Airlie <[email protected]>
Date: Tue, 18 Jun 2013 11:38:10 +1000
Subject: [PATCH] drm: remove vma debug code

This lists vma in /proc and is both crash prone and quite possible horribly racy. Just nuke it I don't think I've used it in years and years.

---
 drivers/gpu/drm/drm_debugfs.c |  3 ---
 drivers/gpu/drm/drm_info.c    | 54 -------------------------------------------
 drivers/gpu/drm/drm_proc.c    |  3 ---
 include/drm/drmP.h            |  4 ----
 4 files changed, 64 deletions(-)
drivers/gpu/drm/drm_debugfs.c +0 -3
index a05087c..595c8c1 100644
@@ -48,9 +48,6 @@ static struct drm_info_list drm_debugfs_list[] = {
 	{"clients", drm_clients_info, 0},
 	{"bufs", drm_bufs_info, 0},
 	{"gem_names", drm_gem_name_info, DRIVER_GEM},
-#if DRM_DEBUG_CODE
-	{"vma", drm_vma_info, 0},
-#endif
 };
 #define DRM_DEBUGFS_ENTRIES ARRAY_SIZE(drm_debugfs_list)
 
drivers/gpu/drm/drm_info.c +0 -54

57 lines not on this page · raw message

drivers/gpu/drm/drm_proc.c +0 -3
index d7f2324..92e9abd 100644
@@ -55,9 +55,6 @@ static const struct drm_info_list drm_proc_list[] = {
 	{"clients", drm_clients_info, 0},
 	{"bufs", drm_bufs_info, 0},
 	{"gem_names", drm_gem_name_info, DRIVER_GEM},
-#if DRM_DEBUG_CODE
-	{"vma", drm_vma_info, 0},
-#endif
 };
 #define DRM_PROC_ENTRIES ARRAY_SIZE(drm_proc_list)
 
include/drm/drmP.h +0 -4
index 63d17ee..849523d 100644
@@ -1600,10 +1600,6 @@ int drm_prime_add_dma_buf(struct drm_device *dev, struct drm_gem_object *obj);
 int drm_prime_lookup_obj(struct drm_device *dev, struct dma_buf *buf,
 			 struct drm_gem_object **obj);
 
-#if DRM_DEBUG_CODE
-extern int drm_vma_info(struct seq_file *m, void *data);
-#endif
-
 				/* Scatter Gather Support (drm_scatter.h) */
 extern void drm_sg_cleanup(struct drm_sg_mem * entry);
 extern int drm_sg_alloc_ioctl(struct drm_device *dev, void *data,
Signature · 1 line
1.8.1.2
author
#2 lore raw
Quoted 9 lines · David Airlie
Quoted 7 lines

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded.
(Note, no X running on that box)

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

How about this, lets just rip it all out.

That's one way to deal with it :)
If no programs use it, then yeah, sure, why not.

Dave

author
#3 lore raw
In reply to David Airlie
Quoted 9 lines · David Airlie
Quoted 7 lines

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded.
(Note, no X running on that box)

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

How about this, lets just rip it all out.

No-one objected, and this is still around in 3.11-rc3 in the same easily oopsable state.. I vote we kill it with fire.

Dave

#4 lore raw
Quoted 6 lines · Dave Jones
Quoted 5 lines · David Airlie
Quoted 5 lines

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded.
(Note, no X running on that box)

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

How about this, lets just rip it all out.

No-one objected, and this is still around in 3.11-rc3 in the same easily oopsable state.. I vote we kill it with fire.

Can we make it burn brighter while at it?

http://cgit.freedesktop.org/~danvet/drm/commit/?h=for-dvdhrm&id=151591c2828e18fde1eb8447874704f3422168b0

Cheers, Daniel

Signature · 3 lines
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
author
#5 lore raw
Quoted 10 lines · Daniel Vetter
Quoted 9 lines · Dave Jones
Quoted 8 lines · David Airlie
Quoted 7 lines

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded.
(Note, no X running on that box)

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

How about this, lets just rip it all out.

No-one objected, and this is still around in 3.11-rc3 in the same easily oopsable state.. I vote we kill it with fire.

Can we make it burn brighter while at it?

http://cgit.freedesktop.org/~danvet/drm/commit/?h=for-dvdhrm&id=151591c2828e18fde1eb8447874704f3422168b0

This went kinda quiet, what's the plan here ?

Dave

maintainer
#6 lore raw
Quoted 13 lines · Dave Jones
Quoted 12 lines · Daniel Vetter
Quoted 11 lines · Dave Jones
Quoted 9 lines · David Airlie
Quoted 7 lines

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded.
(Note, no X running on that box)

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

How about this, lets just rip it all out.

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

No-one objected, and this is still around in 3.11-rc3 in the same easily oopsable state.. I vote we kill it with fire.

Can we make it burn brighter while at it?

http://cgit.freedesktop.org/~danvet/drm/commit/?h=for-dvdhrm&id=151591c2828e18fde1eb8447874704f3422168b0

This went kinda quiet, what's the plan here ?

We nuked it from orbit in drm-next.

Dave.

author
#7 lore raw
Quoted 17 lines · Dave Airlie
Quoted 16 lines · Dave Jones
Quoted 15 lines · Daniel Vetter
Quoted 13 lines · Dave Jones
Quoted 9 lines · David Airlie
Quoted 7 lines

Reading /proc/dri/0/vma causes bad things to happen on a box with nouveau loaded.
(Note, no X running on that box)

Trace below shows trinity, but I can reproduce it with just cat /proc/dri/0/vma

How about this, lets just rip it all out.

No-one objected, and this is still around in 3.11-rc3 in the same easily oopsable state.. I vote we kill it with fire.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

Can we make it burn brighter while at it?

http://cgit.freedesktop.org/~danvet/drm/commit/?h=for-dvdhrm&id=151591c2828e18fde1eb8447874704f3422168b0

This went kinda quiet, what's the plan here ?

We nuked it from orbit in drm-next.

Awesome. Looks like that missed a Cc: -stable tag btw.

Dave