Report split between the beginning and end of the buffer

[matt-auld]

Seems to be pretty damn rare, and only tends to happen when reading OA reports for a prolonged period of time. Reproducible on oa-nextrunning on BDW.

[  534.949540] kernel BUG at drivers/gpu/drm/i915/i915_perf.c:193!
[  534.949558] invalid opcode: 0000 [#1] SMP 
[  534.949573] Modules linked in: rfcomm fuse ccm cmac xt_CHECKSUM ipt_MASQUERADE nf_nat_masquerade_ipv4 tun nf_conntrack_netbios_ns nf_conntrack_broadcast ip6t_rpfilter ip6t_REJECT nf_reject_ipv6 xt_conntrack ebtable_filter ebtable_nat ebtable_broute bridge ebtables ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 ip6table_raw ip6table_security ip6table_mangle ip6table_filter ip6_tables iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_raw iptable_security iptable_mangle bnep joydev hid_sensor_accel_3d hid_sensor_magn_3d hid_sensor_als hid_sensor_rotation hid_sensor_gyro_3d hid_sensor_incl_3d hid_sensor_trigger hid_sensor_iio_common industrialio_triggered_buffer arc4 snd_hda_codec_hdmi hid_multitouch hid_sensor_hub iwlmvm snd_soc_sst_broadwell intel_rapl
[  534.949819]  x86_pkg_temp_thermal coretemp kvm_intel snd_soc_sst_haswell_pcm vfat mac80211 fat kvm iTCO_wdt iTCO_vendor_support snd_soc_sst_ipc snd_soc_sst_dsp snd_soc_rt286 ppdev snd_soc_rl6347a snd_soc_core irqbypass crct10dif_pclmul crc32_pclmul crc32c_intel iwlwifi snd_hda_intel snd_hda_codec cfg80211 snd_hwdep snd_hda_core uvcvideo snd_seq btusb videobuf2_vmalloc snd_seq_device videobuf2_memops btrtl i2c_i801 videobuf2_v4l2 videobuf2_core btbcm snd_pcm v4l2_common btintel bluetooth lpc_ich videodev mei_me mei media snd_timer cdc_acm snd processor_thermal_device intel_soc_dts_iosf iosf_mbi shpchp soundcore acpi_als kfifo_buf industrialio winbond_cir soc_button_array rc_core int3403_thermal tpm_crb dw_dmac tpm parport_pc parport int3402_thermal rfkill_gpio i2c_designware_platform int3400_thermal
[  534.950076]  i2c_designware_core rfkill snd_soc_sst_acpi dw_dmac_core acpi_thermal_rel acpi_pad int340x_thermal_zone nfsd auth_rpcgss nfs_acl lockd grace sunrpc i915 i2c_algo_bit drm_kms_helper drm e1000e serio_raw sdhci_pci ptp sdhci_acpi pps_core sdhci 8021q mmc_core garp stp llc video mrp i2c_hid cdc_mbim cdc_wdm cdc_ncm usbnet
[  534.950190] CPU: 3 PID: 3312 Comm: glxgears Not tainted 4.4.0-drm-intel-matthew-old+ #9
[  534.950213] Hardware name: Intel Corporation Broadwell Client platform/Wilson Beach SDS, BIOS BDW-E2R1.86C.0095.R09.1410300006 10/30/2014
[  534.950246] task: ffff88023c055700 ti: ffff88008cb04000 task.ti: ffff88008cb04000
[  534.950266] RIP: 0010:[<ffffffffa02732e6>]  [<ffffffffa02732e6>] gen8_oa_read+0x186/0x260 [i915]
[  534.950314] RSP: 0018:ffff88008cb07d88  EFLAGS: 00010287
[  534.950329] RAX: 00000000000000c0 RBX: 00000000000f28dd RCX: 0000000000000000
[  534.950349] RDX: 0000000000000100 RSI: 0000000000000000 RDI: 000055d11d540b98
[  534.950369] RBP: ffff88008cb07dd0 R08: 0000000000ffff40 R09: ffff88023ef45a00
[  534.950388] R10: ffffc90004801000 R11: ffff88023c00cb00 R12: 00000000ff000000
[  534.950408] R13: 0000000000000100 R14: 0000000000ffff40 R15: ffff88023d550000
[  534.950428] FS:  00007f15cd36c700(0000) GS:ffff88024dcc0000(0000) knlGS:0000000000000000
[  534.950450] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  534.950466] CR2: 00007f7bd2de7018 CR3: 0000000200cd0000 CR4: 00000000003406e0
[  534.950485] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  534.950505] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  534.950524] Stack:
[  534.950531]  ffff88023d550000 ffff88008cb07df8 ffffc90004801000 ffff88023ef45a00
[  534.950555]  ffff88023ef45a00 ffff880090a68f00 ffff88023d55a028 ffff88023d550000
[  534.950580]  000055d11d540880 ffff88008cb07de0 ffffffffa0272582 ffff88008cb07e40
[  534.950604] Call Trace:
[  534.950629]  [<ffffffffa0272582>] i915_oa_read+0x12/0x20 [i915]
[  534.950661]  [<ffffffffa0272d2c>] i915_perf_read+0x7c/0xf0 [i915]
[  534.950681]  [<ffffffff81222567>] __vfs_read+0x37/0x100
[  534.950697]  [<ffffffff81328590>] ? security_file_permission+0xa0/0xc0
[  534.950716]  [<ffffffff81222e96>] vfs_read+0x86/0x130
[  534.950731]  [<ffffffff81223be5>] SyS_read+0x55/0xc0
[  534.950746]  [<ffffffff81235e63>] ? SyS_ioctl+0x63/0x90
[  534.950763]  [<ffffffff817d1cee>] entry_SYSCALL_64_fastpath+0x12/0x71
[  534.950781] Code: 00 25 ff ff ff 00 74 25 44 39 e8 72 20 45 89 f0 b8 00 00 00 01 45 29 ec 41 81 e0 ff ff ff 00 44 29 c0 44 39 e8 0f 83 60 ff ff ff <0f> 0b 45 89 f5 41 8b 97 38 a1 00 00 4c 8b 75 b8 44 01 ea 4c 89 
[  534.950896] RIP  [<ffffffffa02732e6>] gen8_oa_read+0x186/0x260 [i915]
[  534.950929]  RSP <ffff88008cb07d88>

[rib]

With the number of i915 perf changes since this was reported, I'm closing this for now while I at least haven't seen anything like this on SKL. Might need to re-open if this crops up a gain with more BDW testing.