EDMA API is broken #72
Assignees
Labels
Comments
|
The DMA driver in 4.1 uses the DMA Engine API. The DMA Engine is partially documented in the kernel, but if you are looking for a porting example, you can take a look at logi-kernel. |
|
As a matter of fact, the given example from TI does not work in Kernel 4.4+. The Linux DMA Engine API does work for 4.1-ti kernels but the performance is worse compared to any other solution. However, the same code does not work for the default 4.4-bone kernels. Does anybody know of an example how to access the EDMA driver directly and not via the Linux DMA Engine API in kernel 4.4+? |
RobertCNelson
pushed a commit
that referenced
this issue
Oct 25, 2017
kvmppc_gpa_to_ua() accesses KVM memory slot array via
srcu_dereference_check() and this produces warnings from RCU like below.
This extends the existing srcu_read_lock/unlock to cover that
kvmppc_gpa_to_ua() as well.
We did not hit this before as this lock is not needed for the realmode
handlers and hash guests would use the realmode path all the time;
however the radix guests are always redirected to the virtual mode
handlers and hence the warning.
[ 68.253798] ./include/linux/kvm_host.h:575 suspicious rcu_dereference_check() usage!
[ 68.253799]
other info that might help us debug this:
[ 68.253802]
rcu_scheduler_active = 2, debug_locks = 1
[ 68.253804] 1 lock held by qemu-system-ppc/6413:
[ 68.253806] #0: (&vcpu->mutex){+.+.}, at: [<c00800000e3c22f4>] vcpu_load+0x3c/0xc0 [kvm]
[ 68.253826]
stack backtrace:
[ 68.253830] CPU: 92 PID: 6413 Comm: qemu-system-ppc Tainted: G W 4.14.0-rc3-00553-g432dcba58e9c-dirty #72
[ 68.253833] Call Trace:
[ 68.253839] [c000000fd3d9f790] [c000000000b7fcc8] dump_stack+0xe8/0x160 (unreliable)
[ 68.253845] [c000000fd3d9f7d0] [c0000000001924c0] lockdep_rcu_suspicious+0x110/0x180
[ 68.253851] [c000000fd3d9f850] [c0000000000e825c] kvmppc_gpa_to_ua+0x26c/0x2b0
[ 68.253858] [c000000fd3d9f8b0] [c00800000e3e1984] kvmppc_h_put_tce+0x12c/0x2a0 [kvm]
Fixes: 121f80b ("KVM: PPC: VFIO: Add in-kernel acceleration for VFIO")
Cc: [email protected] # v4.12+
Signed-off-by: Alexey Kardashevskiy <[email protected]>
Signed-off-by: Paul Mackerras <[email protected]>
RobertCNelson
pushed a commit
that referenced
this issue
Aug 20, 2018
commit 506a66f upstream Dave Hansen reported, that it's outright dangerous to keep SMT siblings disabled completely so they are stuck in the BIOS and wait for SIPI. The reason is that Machine Check Exceptions are broadcasted to siblings and the soft disabled sibling has CR4.MCE = 0. If a MCE is delivered to a logical core with CR4.MCE = 0, it asserts IERR#, which shuts down or reboots the machine. The MCE chapter in the SDM contains the following blurb: Because the logical processors within a physical package are tightly coupled with respect to shared hardware resources, both logical processors are notified of machine check errors that occur within a given physical processor. If machine-check exceptions are enabled when a fatal error is reported, all the logical processors within a physical package are dispatched to the machine-check exception handler. If machine-check exceptions are disabled, the logical processors enter the shutdown state and assert the IERR# signal. When enabling machine-check exceptions, the MCE flag in control register CR4 should be set for each logical processor. Reverting the commit which ignores siblings at enumeration time solves only half of the problem. The core cpuhotplug logic needs to be adjusted as well. This thoughtful engineered mechanism also turns the boot process on all Intel HT enabled systems into a MCE lottery. MCE is enabled on the boot CPU before the secondary CPUs are brought up. Depending on the number of physical cores the window in which this situation can happen is smaller or larger. On a HSW-EX it's about 750ms: MCE is enabled on the boot CPU: [ 0.244017] mce: CPU supports 22 MCE banks The corresponding sibling #72 boots: [ 1.008005] .... node #0, CPUs: #72 That means if an MCE hits on physical core 0 (logical CPUs 0 and 72) between these two points the machine is going to shutdown. At least it's a known safe state. It's obvious that the early boot can be hit by an MCE as well and then runs into the same situation because MCEs are not yet enabled on the boot CPU. But after enabling them on the boot CPU, it does not make any sense to prevent the kernel from recovering. Adjust the nosmt kernel parameter documentation as well. Reverts: 2207def ("x86/apic: Ignore secondary threads if nosmt=force") Reported-by: Dave Hansen <[email protected]> Signed-off-by: Thomas Gleixner <[email protected]> Tested-by: Tony Luck <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
RobertCNelson
pushed a commit
that referenced
this issue
Oct 22, 2018
commit 506a66f upstream Dave Hansen reported, that it's outright dangerous to keep SMT siblings disabled completely so they are stuck in the BIOS and wait for SIPI. The reason is that Machine Check Exceptions are broadcasted to siblings and the soft disabled sibling has CR4.MCE = 0. If a MCE is delivered to a logical core with CR4.MCE = 0, it asserts IERR#, which shuts down or reboots the machine. The MCE chapter in the SDM contains the following blurb: Because the logical processors within a physical package are tightly coupled with respect to shared hardware resources, both logical processors are notified of machine check errors that occur within a given physical processor. If machine-check exceptions are enabled when a fatal error is reported, all the logical processors within a physical package are dispatched to the machine-check exception handler. If machine-check exceptions are disabled, the logical processors enter the shutdown state and assert the IERR# signal. When enabling machine-check exceptions, the MCE flag in control register CR4 should be set for each logical processor. Reverting the commit which ignores siblings at enumeration time solves only half of the problem. The core cpuhotplug logic needs to be adjusted as well. This thoughtful engineered mechanism also turns the boot process on all Intel HT enabled systems into a MCE lottery. MCE is enabled on the boot CPU before the secondary CPUs are brought up. Depending on the number of physical cores the window in which this situation can happen is smaller or larger. On a HSW-EX it's about 750ms: MCE is enabled on the boot CPU: [ 0.244017] mce: CPU supports 22 MCE banks The corresponding sibling #72 boots: [ 1.008005] .... node #0, CPUs: #72 That means if an MCE hits on physical core 0 (logical CPUs 0 and 72) between these two points the machine is going to shutdown. At least it's a known safe state. It's obvious that the early boot can be hit by an MCE as well and then runs into the same situation because MCEs are not yet enabled on the boot CPU. But after enabling them on the boot CPU, it does not make any sense to prevent the kernel from recovering. Adjust the nosmt kernel parameter documentation as well. Reverts: 2207def ("x86/apic: Ignore secondary threads if nosmt=force") Reported-by: Dave Hansen <[email protected]> Signed-off-by: Thomas Gleixner <[email protected]> Tested-by: Tony Luck <[email protected]> Signed-off-by: David Woodhouse <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
crow-misia
pushed a commit
to crow-misia/linux
that referenced
this issue
Mar 24, 2019
commit baef1c9 upstream. Using the batch API from the interconnect driver sometimes leads to a KASAN error due to an access to freed memory. This is easier to trigger with threadirqs on the kernel commandline. BUG: KASAN: use-after-free in rpmh_tx_done+0x114/0x12c Read of size 1 at addr fffffff51414ad84 by task irq/110-apps_rs/57 CPU: 0 PID: 57 Comm: irq/110-apps_rs Tainted: G W 4.19.10 beagleboard#72 Call trace: dump_backtrace+0x0/0x2f8 show_stack+0x20/0x2c __dump_stack+0x20/0x28 dump_stack+0xcc/0x10c print_address_description+0x74/0x240 kasan_report+0x250/0x26c __asan_report_load1_noabort+0x20/0x2c rpmh_tx_done+0x114/0x12c tcs_tx_done+0x450/0x768 irq_forced_thread_fn+0x58/0x9c irq_thread+0x120/0x1dc kthread+0x248/0x260 ret_from_fork+0x10/0x18 Allocated by task 385: kasan_kmalloc+0xac/0x148 __kmalloc+0x170/0x1e4 rpmh_write_batch+0x174/0x540 qcom_icc_set+0x8dc/0x9ac icc_set+0x288/0x2e8 a6xx_gmu_stop+0x320/0x3c0 a6xx_pm_suspend+0x108/0x124 adreno_suspend+0x50/0x60 pm_generic_runtime_suspend+0x60/0x78 __rpm_callback+0x214/0x32c rpm_callback+0x54/0x184 rpm_suspend+0x3f8/0xa90 pm_runtime_work+0xb4/0x178 process_one_work+0x544/0xbc0 worker_thread+0x514/0x7d0 kthread+0x248/0x260 ret_from_fork+0x10/0x18 Freed by task 385: __kasan_slab_free+0x12c/0x1e0 kasan_slab_free+0x10/0x1c kfree+0x134/0x588 rpmh_write_batch+0x49c/0x540 qcom_icc_set+0x8dc/0x9ac icc_set+0x288/0x2e8 a6xx_gmu_stop+0x320/0x3c0 a6xx_pm_suspend+0x108/0x124 adreno_suspend+0x50/0x60 cr50_spi spi5.0: SPI transfer timed out pm_generic_runtime_suspend+0x60/0x78 __rpm_callback+0x214/0x32c rpm_callback+0x54/0x184 rpm_suspend+0x3f8/0xa90 pm_runtime_work+0xb4/0x178 process_one_work+0x544/0xbc0 worker_thread+0x514/0x7d0 kthread+0x248/0x260 ret_from_fork+0x10/0x18 The buggy address belongs to the object at fffffff51414ac80 which belongs to the cache kmalloc-512 of size 512 The buggy address is located 260 bytes inside of 512-byte region [fffffff51414ac80, fffffff51414ae80) The buggy address belongs to the page: page:ffffffbfd4505200 count:1 mapcount:0 mapping:fffffff51e00c680 index:0x0 compound_mapcount: 0 flags: 0x4000000000008100(slab|head) raw: 4000000000008100 ffffffbfd4529008 ffffffbfd44f9208 fffffff51e00c680 raw: 0000000000000000 0000000000200020 00000001ffffffff 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: fffffff51414ac80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fffffff51414ad00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb >fffffff51414ad80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ fffffff51414ae00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fffffff51414ae80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc The batch API sets the same completion for each rpmh message that's sent and then loops through all the messages and waits for that single completion declared on the stack to be completed before returning from the function and freeing the message structures. Unfortunately, some messages may still be in process and 'stuck' in the TCS. At some later point, the tcs_tx_done() interrupt will run and try to process messages that have already been freed at the end of rpmh_write_batch(). This will in turn access the 'needs_free' member of the rpmh_request structure and cause KASAN to complain. Furthermore, if there's a message that's completed in rpmh_tx_done() and freed immediately after the complete() call is made we'll be racing with potentially freed memory when accessing the 'needs_free' member: CPU0 CPU1 ---- ---- rpmh_tx_done() complete(&compl) wait_for_completion(&compl) kfree(rpm_msg) if (rpm_msg->needs_free) <KASAN warning splat> Let's fix this by allocating a chunk of completions for each message and waiting for all of them to be completed before returning from the batch API. Alternatively, we could wait for the last message in the batch, but that may be a more complicated change because it looks like tcs_tx_done() just iterates through the indices of the queue and completes each message instead of tracking the last inserted message and completing that first. Fixes: c8790cb ("drivers: qcom: rpmh: add support for batch RPMH request") Cc: Lina Iyer <[email protected]> Cc: "Raju P.L.S.S.S.N" <[email protected]> Cc: Matthias Kaehlcke <[email protected]> Cc: Evan Green <[email protected]> Cc: [email protected] Reviewed-by: Lina Iyer <[email protected]> Reviewed-by: Evan Green <[email protected]> Signed-off-by: Stephen Boyd <[email protected]> Signed-off-by: Bjorn Andersson <[email protected]> Signed-off-by: Andy Gross <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
RobertCNelson
pushed a commit
that referenced
this issue
Sep 25, 2019
[ Upstream commit 23377c2 ] When the device is disconnected while passing traffic it is possible to receive out of order urbs causing a memory leak since the skb linked to the current tx urb is not removed. Fix the issue deallocating the skb cleaning up the tx ring. Moreover this patch fixes the following kernel warning [ 57.480771] usb 1-1: USB disconnect, device number 2 [ 57.483451] ------------[ cut here ]------------ [ 57.483462] TX urb mismatch [ 57.483481] WARNING: CPU: 1 PID: 32 at drivers/net/wireless/mediatek/mt7601u/dma.c:245 mt7601u_complete_tx+0x165/00 [ 57.483483] Modules linked in: [ 57.483496] CPU: 1 PID: 32 Comm: kworker/1:1 Not tainted 5.2.0-rc1+ #72 [ 57.483498] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.12.0-2.fc30 04/01/2014 [ 57.483502] Workqueue: usb_hub_wq hub_event [ 57.483507] RIP: 0010:mt7601u_complete_tx+0x165/0x1e0 [ 57.483510] Code: 8b b5 10 04 00 00 8b 8d 14 04 00 00 eb 8b 80 3d b1 cb e1 00 00 75 9e 48 c7 c7 a4 ea 05 82 c6 05 f [ 57.483513] RSP: 0000:ffffc900000a0d28 EFLAGS: 00010092 [ 57.483516] RAX: 000000000000000f RBX: ffff88802c0a62c0 RCX: ffffc900000a0c2c [ 57.483518] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffffffff810a8371 [ 57.483520] RBP: ffff88803ced6858 R08: 0000000000000000 R09: 0000000000000001 [ 57.483540] R10: 0000000000000002 R11: 0000000000000000 R12: 0000000000000046 [ 57.483542] R13: ffff88802c0a6c88 R14: ffff88803baab540 R15: ffff88803a0cc078 [ 57.483548] FS: 0000000000000000(0000) GS:ffff88803eb00000(0000) knlGS:0000000000000000 [ 57.483550] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 57.483552] CR2: 000055e7f6780100 CR3: 0000000028c86000 CR4: 00000000000006a0 [ 57.483554] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 57.483556] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 57.483559] Call Trace: [ 57.483561] <IRQ> [ 57.483565] __usb_hcd_giveback_urb+0x77/0xe0 [ 57.483570] xhci_giveback_urb_in_irq.isra.0+0x8b/0x140 [ 57.483574] handle_cmd_completion+0xf5b/0x12c0 [ 57.483577] xhci_irq+0x1f6/0x1810 [ 57.483581] ? lockdep_hardirqs_on+0x9e/0x180 [ 57.483584] ? _raw_spin_unlock_irq+0x24/0x30 [ 57.483588] __handle_irq_event_percpu+0x3a/0x260 [ 57.483592] handle_irq_event_percpu+0x1c/0x60 [ 57.483595] handle_irq_event+0x2f/0x4c [ 57.483599] handle_edge_irq+0x7e/0x1a0 [ 57.483603] handle_irq+0x17/0x20 [ 57.483607] do_IRQ+0x54/0x110 [ 57.483610] common_interrupt+0xf/0xf [ 57.483612] </IRQ> Acked-by: Jakub Kicinski <[email protected]> Signed-off-by: Lorenzo Bianconi <[email protected]> Signed-off-by: Kalle Valo <[email protected]> Signed-off-by: Sasha Levin <[email protected]>
|
@QuickJack @jcdevel @ViktorButskikh Have you tried the 4.19 kernel on our current image? |
|
You can find an overview of the kernels that I have tested at https://github.com/fpga-logi/logi-kernel. I guess that upgrading the kernel to a newer version might require an upgrade of the driver code as well. |
Is the driver that uses EDMA inside here? It looks like this is the kernel module? |
|
The code contains support for different hardware platforms (Logibone R1, Logibone RA2). The main part is in https://github.com/fpga-logi/logi-kernel/blob/master/beaglebone-black/common/main_dm.c and https://github.com/fpga-logi/logi-kernel/blob/master/beaglebone-black/common/main_dma.c. The latter file contains the DMA support which is currently used by default. The actual DMA functionality is implemented in https://github.com/fpga-logi/logi-kernel/blob/master/beaglebone-black/common/logi_dma.c. |
|
@ViktorButskikh are you still looking for a solution to this issue? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hello.
After migration from 3.8.13 kernel to 4.1.18 on my BBB I have found that compilation of my custom kernel driver which using EDMA API is impossible.
My driver based on this classical tutorial http://processors.wiki.ti.com/index.php/EDMA_sample_test_application. But now in file linux/platform_data/edma.h absent basic function for work with EDMA like as edma_start, edma_stop, edma_alloc_channel ...
it's happened after 0f30e8c commit.
Which guide should I use for porting my application from old kenel to new ?
Thanks
The text was updated successfully, but these errors were encountered: