Message-ID: <20190221143803.GD2549@minyard.net>
Date:   Thu, 21 Feb 2019 08:38:03 -0600
From:   Corey Minyard <minyard@....org>
To:     Yang Yingliang <yangyingliang@...wei.com>
Cc:     cminyard@...sta.com, arnd@...db.de,
        openipmi-developer@...ts.sourceforge.net,
        linux-kernel@...r.kernel.org, honglei5@...wei.com
Subject: Re: [PATCH] ipmi_si: fix oops when loading ipmi_si driver

On Thu, Feb 21, 2019 at 02:35:41PM +0800, Yang Yingliang wrote:
> When we excute the following commands, we got oops
> modprobe ipmi_si ports=0xffc0e3 type=bt

Hmm, I have been contemplating pulling out the hardcode interface from
IPMI, I didn't know anyone was using it.  It obviously hasn't been
tested in a while.

The trouble here is that ipmi_si_hardcode_find_bmc() is called before
ipmi_si_platform_init(), which does the a similar thing as the code you
added, except it calls platform_driver_register().  Your change would
actually break other IPMI platform driver devices because
platform_driver_register() does other necessary things.

The theory here is that you want the hard-coded devices to override
anything else, because maybe the SMBIOS tables got the interrupt or
register spacing wrong or something like that.

The right thing to do here is to add the hard-coded devices with
with platform_device_add(), and remove them on module unload.  But
that breaks the theory, at module probe time the other devices
will already be registered with ACPI, SMBIOS, or OF, so the hard
coded devices will not come first.

My thought right now is to do the platform_device_add() with the
hard-coded devices, then add something at the top of ipmi_si_add_smi()
that will reject any non-hard-coded device that has a hard-coded
device registered at the same address/address space.

I'll attempt a patch like that.

-corey

> 
> [  503.305487] ipmi_si: IPMI System Interface driver
> [  503.305489] ipmi_hardcode: probing via hardcoded address
> [  503.305491] ipmi_si: Adding hardcoded-specified bt state machine
> [  503.305494] ipmi_si: Trying hardcoded-specified bt state machine at i/o address 0xffc0e3, slave address 0x0, irq 0
> [  503.337865] ipmi_si ipmi_si.0: bt cap response too short: 3
> [  503.337867] ipmi_si ipmi_si.0: using default values
> [  503.337869] ipmi_si ipmi_si.0: req2rsp=5 secs retries=2
> [  503.421948] ipmi_si ipmi_si.0: IPMI message handler: Found new BMC (man_id: 0x0007db, prod_id: 0x0001, dev_id: 0x01)
> [  503.665959] ipmi_si ipmi_si.0: IPMI bt interface initialized
> [  512.188061] ipmi_si: module verification failed: signature and/or required key missing - tainting kernel
> [  512.190063] ipmi_si: IPMI System Interface driver
> [  512.190065] ipmi_hardcode: probing via hardcoded address
> [  512.190067] ipmi_si: Adding hardcoded-specified bt state machine
> [  512.190070] ipmi_si: Trying hardcoded-specified bt state machine at i/o address 0xffc0e3, slave address 0x0, irq 0
> [  512.201867] ipmi_si ipmi_si.0: bt cap response too short: 3
> [  512.201869] ipmi_si ipmi_si.0: using default values
> [  512.201870] ipmi_si ipmi_si.0: req2rsp=5 secs retries=2
> [  512.269898] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000030
> [  512.269899] Mem abort info:
> [  512.269900]   ESR = 0x96000006
> [  512.269902]   Exception class = DABT (current EL), IL = 32 bits
> [  512.269903]   SET = 0, FnV = 0
> [  512.269904]   EA = 0, S1PTW = 0
> [  512.269905] Data abort info:
> [  512.269906]   ISV = 0, ISS = 0x00000006
> [  512.269908]   CM = 0, WnR = 0
> [  512.269910] user pgtable: 4k pages, 48-bit VAs, pgdp = 000000003f829971
> [  512.269912] [0000000000000030] pgd=0000005f1e7de003, pud=0000005f69728003, pmd=0000000000000000
> [  512.269916] Internal error: Oops: 96000006 [#1] SMP
> [  512.274923] Modules linked in: ipmi_si(E+) nls_utf8 isofs dm_mirror dm_region_hash dm_log dm_mod aes_ce_blk crypto_simd cryptd aes_ce_cipher ghash_ce sha2_ce sha256_arm64 sha1_ce ses hibmc_drm enclosure hisi_sas_v2_hw hisi_sas_main sg sbsa_gwdt ip_tables marvell ixgbe hns_dsaf hns_enet_drv ipmi_devintf mpt3sas ipmi_msghandler hns_mdio hnae mdio [last unloaded: ipmi_si]
> [  512.308100] CPU: 27 PID: 13691 Comm: modprobe Kdump: loaded Tainted: G            E     5.0.0-rc7+ #273
> [  512.317456] Hardware name: Huawei TaiShan 2280 /BC11SPCD, BIOS 1.37 11/21/2017
> [  512.324781] pstate: a0000005 (NzCv daif -PAN -UAO)
> [  512.329643] pc : sysfs_do_create_link_sd.isra.0+0x5c/0x110
> [  512.335204] lr : sysfs_do_create_link_sd.isra.0+0x50/0x110
> [  512.340763] sp : ffff000012a8b8b0
> [  512.344118] x29: ffff000012a8b8b0 x28: ffff805f69bf8818
> [  512.349505] x27: 0000000000000000 x26: 0000000000000000
> [  512.354891] x25: ffff000011840000 x24: ffff805f69248ee0
> [  512.360277] x23: 0000000000000001 x22: ffff000011b0d000
> [  512.365662] x21: ffff00001116e3e0 x20: ffff805f76db7cb0
> [  512.371047] x19: 0000000000000030 x18: ffffffffffffffff
> [  512.376433] x17: 0000000000000000 x16: 0000000000000000
> [  512.381818] x15: ffff00001176d708 x14: 5244006d726f6674
> [  512.387204] x13: 0000000000000040 x12: 0000000000000008
> [  512.392589] x11: 0101010101010101 x10: 7f7f7f7f7f7f7f7f
> [  512.397974] x9 : 716e6573606aff2f x8 : 7f7f7f7f7f7f7f7f
> [  512.403360] x7 : 2d68725e686c6f68 x6 : 0080000000000000
> [  512.408745] x5 : 0000000000000000 x4 : ffff805f08e2d440
> [  512.414131] x3 : ffff000011b0df60 x2 : 0000000000000001
> [  512.419516] x1 : 0000000000000000 x0 : 0000000000000000
> [  512.424902] Process insmod (pid: 13691, stack limit = 0x0000000007e6d9d7)
> [  512.431784] Call trace:
> [  512.434260]  sysfs_do_create_link_sd.isra.0+0x5c/0x110
> [  512.439468]  sysfs_create_link+0x40/0x68
> [  512.443446]  driver_sysfs_add+0x88/0xb8
> [  512.447332]  device_bind_driver+0x20/0x70
> [  512.451392]  __device_attach+0xa0/0x170
> [  512.459409]  device_initial_probe+0x24/0x30
> [  512.467809]  bus_probe_device+0xa0/0xa8
> [  512.475844]  device_add+0x494/0x620
> [  512.483441]  platform_device_add+0x118/0x2a0
> [  512.491638]  try_smi_init+0x6c0/0x11c8 [ipmi_si]
> [  512.500161]  init_ipmi_si+0x13c/0x1f0 [ipmi_si]
> [  512.508342]  do_one_initcall+0x54/0x1f0
> [  512.515668]  do_init_module+0x64/0x1e4
> [  512.522810]  load_module+0x13fc/0x14f8
> [  512.529816]  __se_sys_finit_module+0xa0/0x100
> [  512.537398]  __arm64_sys_finit_module+0x24/0x30
> [  512.545045]  el0_svc_common+0x120/0x148
> [  512.552029]  el0_svc_handler+0x38/0x78
> [  512.558918]  el0_svc+0x8/0xc
> [  512.564960] Code: 942513e5 d503201f d503201f 35000400 (f9400273)
> [  512.574273] ---[ end trace 2a5abdfb7a1edf93 ]---
> [  512.582123] Kernel panic - not syncing: Fatal exception
> [  512.590600] SMP: stopping secondary CPUs
> [  512.597934] Kernel Offset: disabled
> [  512.604635] CPU features: 0x002,21006008
> [  512.611889] Memory Limit: none
> [  512.621495] Starting crashdump kernel...
> [  512.628775] Bye!
> 
> Because the target_kobj pointer is null in sysfs_do_create_link_sd().
> The pointer is assigned by new_smi->io.dev->driver->p.
> 
> If ipmi_si driver is not registered, new_smi->io.dev->driver->p is null
> and it will be used when try_smi_init() calling platform_device_add().
> Fix this oops by registering the driver before calling platform_device_add().
> 
> Fixes: 9d70029edbbf ("ipmi_si: Move platform device handling to another file")
> 
> Reported-by: Hong Lei <honglei5@...wei.com>
> Signed-off-by: Yang Yingliang <yangyingliang@...wei.com>
> ---
>  drivers/char/ipmi/ipmi_si_hardcode.c | 7 +++++++
>  1 file changed, 7 insertions(+)
> 
> diff --git a/drivers/char/ipmi/ipmi_si_hardcode.c b/drivers/char/ipmi/ipmi_si_hardcode.c
> index 4876428..2aa1ea9 100644
> --- a/drivers/char/ipmi/ipmi_si_hardcode.c
> +++ b/drivers/char/ipmi/ipmi_si_hardcode.c
> @@ -3,6 +3,7 @@
>  #define pr_fmt(fmt) "ipmi_hardcode: " fmt
>  
>  #include <linux/moduleparam.h>
> +#include <linux/platform_device.h>
>  #include "ipmi_si.h"
>  
>  /*
> @@ -144,5 +145,11 @@ int ipmi_si_hardcode_find_bmc(void)
>  
>  		ret = ipmi_si_add_smi(&io);
>  	}
> +
> +	if (!ret) {
> +		ipmi_platform_driver.driver.bus = &platform_bus_type;
> +		ret = driver_register(&ipmi_platform_driver.driver);
> +	}
> +
>  	return ret;
>  }
> -- 
> 1.8.3.1
> 
> 

Powered by blists - more mailing lists