[Intel-gfx] [PATCH] drm/i915: Wrap the preallocated BIOS framebuffer and preserve for KMS fbcon v5

Jesse Barnes jbarnes at virtuousgeek.org
Thu Nov 7 20:56:33 CET 2013


Note that with this patch I see this on boot when the preallocated
objects are scanned:

[    3.951078] WARNING: CPU: 1 PID: 398 at lib/list_debug.c:36 __list_add+0x8a/0
xc0()
[    3.951081] list_add double add: new=ffff88003fa42ba8, prev=ffff88003f9903c0,
 next=ffff88003fa42ba8.
[    3.951083] Modules linked in: coretemp snd_hda_intel(+) kvm_intel kvm vfat f
at i915(OF+) snd_hda_codec snd_hwdep crc32_pclmul snd_seq crc32c_intel snd_seq_d
evice snd_pcm ghash_clmulni_intel microcode snd_page_alloc snd_timer snd asix se
rio_raw soundcore usbnet shpchp mii i2c_algo_bit drm_kms_helper(OF) drm(OF) i2c_
core sdhci_acpi video nfsd sdhci mmc_core acpi_cpufreq auth_rpcgss nfs_acl lockd
 sunrpc usb_storage
[    3.951122] CPU: 1 PID: 398 Comm: systemd-udevd Tainted: GF       W  O 3.12.0+ #8
[    3.951125] Hardware name: Intel Corp. VALLEYVIEW B3 PLATFORM/NOTEBOOK, BIOS BBAYCRB1.X64.0060.R40.1309121618 BBAY_X64_R_V60_40 09/12/2013
[    3.951127]  0000000000000009 ffff88003ffd3790 ffffffff81640eb8 ffff88003ffd37d8
[    3.951132]  ffff88003ffd37c8 ffffffff810680dd ffff88003fa42ba8 ffff88003fa42ba8
[    3.951136]  ffff88003f9903c0 ffff8800752ce870 ffff88003f990300 ffff88003ffd3828
[    3.951140] Call Trace:
[    3.951148]  [<ffffffff81640eb8>] dump_stack+0x45/0x56
[    3.951154]  [<ffffffff810680dd>] warn_slowpath_common+0x7d/0xa0
[    3.951158]  [<ffffffff8106814c>] warn_slowpath_fmt+0x4c/0x50
[    3.951178]  [<ffffffffa01313f5>] ? drm_mm_reserve_node+0xe5/0x170 [drm]
[    3.951199]  [<ffffffff8130e6ea>] __list_add+0x8a/0xc0
[    3.951230]  [<ffffffffa02863a0>] i915_gem_setup_global_gtt+0xc0/0x280 [i915]
[    3.951256]  [<ffffffffa02869a4>] i915_gem_init_global_gtt+0x444/0x980 [i915]
[    3.951280]  [<ffffffffa0290f38>] ? i9xx_get_fb_config+0x268/0x370 [i915]
[    3.951304]  [<ffffffffa027c2a2>] i915_gem_init+0x42/0x180 [i915]
[    3.951324]  [<ffffffffa0266080>] i915_driver_load+0xc50/0xf10 [i915]
...

So something is still off with the block management of stolen objects
in this patch...

-- 
Jesse Barnes, Intel Open Source Technology Center



More information about the Intel-gfx mailing list