April 2019 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

How to make /dev/ttyACM0 (and friends) exclusive?
by Jeffrey Walton Oct. 11, 2020

Oct. 11, 2020

Hi Everyone, I'm trying to trackdown a problem using my dialup modem. I have a program that opens the modem and watches caller id's. It flashes the hook when a telemarketer calls. It works well until... When another program opens the modem then my program starts reading intermittent responses intended for the other program. I cannot seem to open the device in exclusive mode. The current open is (I also tried with O_EXCL): int modem = open(device_path, O_RDWR | O_NOCTTY | O_SYNC); I'm not sure if it because udev configures the device incorrectly or the kernel driver is configured incorrectly. Or maybe it is impossible to do. Does anyone know how I can make the device exclusive? Thanks in advance.

6 12

debugfs question ...
by rdq＠metamail.co May 1, 2019

May 1, 2019

Greetings, My I2C sensor driver has a debugfs entry for development purposes. Everything works fine with the exception of the read operation. When 'cat' is used, the read operation is called repeatedly and indefinitely. If the read() is changed to return 0 then, as expected, nothing is displayed. The pattern for the implementation is (AFAICT) right out of the book (shown below). What am I missing? Any thoughts much appreciated. TAIA. RDQ static ssize_t sc031gs_reg_read_file(struct file *file, char __user *user_buf, size_t count, loff_t *ppos) { char *buf = 0; ssize_t total = 0; struct sc031gs_dev *sensor = file->private_data; if (!sensor) return -EINVAL; if (*ppos < 0 || !count) return -EINVAL; buf = kmalloc(count, GFP_KERNEL); if (!buf) return -ENOMEM; total = snprintf(buf,count,"Hello world\n"); if (total >= 0) { if (copy_to_user(user_buf, buf, total)) { kfree(buf); return -EFAULT; } *ppos += total; } kfree(buf); return total; }

3 3

vmscan.c: Reclaim unevictable pages.
by Pankaj Suryawanshi April 28, 2019

April 28, 2019

Hello , shrink_page_list() returns , number of pages reclaimed, when pages is unevictable it returns VM_BUG_ON_PAGE(PageLRU(page) || PageUnevicatble(page),page); We can add the unevictable pages in reclaim list in shrink_page_list(), return total number of reclaim pages including unevictable pages, let the caller handle unevictable pages. I think the problem is shrink_page_list is awkard. If page is unevictable it goto activate_locked->keep_locked->keep lables, keep lable list_add the unevictable pages and throw the VM_BUG instead of passing it to caller while it relies on caller for non-reclaimed-non-unevictable page's putback. I think we can make it consistent so that shrink_page_list could return non-reclaimed pages via page_list and caller can handle it. As an advance, it could try to migrate mlocked pages without retrial. Below is the issue i observed of CMA_ALLOC of large size buffer : (Kernel version - 4.14.65 With Android Pie. [ 24.718792] page dumped because: VM_BUG_ON_PAGE(PageLRU(page) || PageUnevictable(page)) [ 24.726949] page->mem_cgroup:bd008c00 [ 24.730693] ------------[ cut here ]------------ [ 24.735304] kernel BUG at mm/vmscan.c:1350! [ 24.739478] Internal error: Oops - BUG: 0 [#1] PREEMPT SMP ARM Below is the patch which solved this issue : diff --git a/mm/vmscan.c b/mm/vmscan.c index be56e2e..12ac353 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -998,7 +998,7 @@ static unsigned long shrink_page_list(struct list_head *page_list, sc->nr_scanned++; if (unlikely(!page_evictable(page))) - goto activate_locked; + goto cull_mlocked; if (!sc->may_unmap && page_mapped(page)) goto keep_locked; @@ -1331,7 +1331,12 @@ static unsigned long shrink_page_list(struct list_head *page_list, } else list_add(&page->lru, &free_pages); continue; - +cull_mlocked: + if (PageSwapCache(page)) + try_to_free_swap(page); + unlock_page(page); + list_add(&page->lru, &ret_pages); + continue; activate_locked: /* Not a candidate for swapping, so reclaim swap space. */ if (PageSwapCache(page) && (mem_cgroup_swap_full(page) || It fixes the below issue. 1. Large size buffer allocation using cma_alloc successful with unevictable pages. cma_alloc of current kernel will fail due to unevictable page Please let me know if anything i am missing. Regards, Pankaj

2 2

Best practice to upgrade kernel
by Xiaolong Wang April 26, 2019

April 26, 2019

Hi all, I have a newbie question. I’m using a 3.18.66 kernel branch from Qualcomm. Over the year I’ve applied some of patched here and there manually. But It seems Qualcomm does not patch lot of newly released vulnerabilities. I would love to upgrade to the latest 3.18.138 (not 4.x) What is the best practice here? Thanks -Dan

3 2

Partition linux page cache for each VM
by Chirag Chauhan April 26, 2019

April 26, 2019

Hey, I need to modify pagecache such that each VM gets their own LRU list and eviction is done from that list. I need to verify whether I'm on the right path, so can anyone please verify my approach. The first step is to identify which VM is making the call to allocate a page. As VMs are stored as a file, I can identify each VM with their filename, right? For this, I'm trying to get the filename associated with an address_space object when add_to_page_cache_lru() is called in filemap.c >From address_space object, I can get inode and therefore dentry. I'm not able to figure out how to get the name of a file. And can you please provide brief overview of what the overall approach should be? Any comments would help. Thanks. Chirag

3 4

__clksrc_of_table and "clocksource_probe: no matching clocksources found"
by Guilherme Costa April 23, 2019

April 23, 2019

Hi, I'm working with kernel 4.9.51 LTSI on a custom board with an AM3352. The system seems to be working just fine, but I'm getting the following message on boot: "clocksource_probe: no matching clocksources found" I did some snooping around and this message is printed by clocksource_probe(), called during time_init(). This is the function code: void __init clocksource_probe(void) { struct device_node *np; const struct of_device_id *match; of_init_fn_1_ret init_func_ret; unsigned clocksources = 0; int ret; for_each_matching_node_and_match(np, __clksrc_of_table, &match) { if (!of_device_is_available(np)) continue; init_func_ret = match->data; ret = init_func_ret(np); if (ret) { pr_err("Failed to initialize '%s': %d", of_node_full_name(np), ret); continue; } clocksources++; } clocksources += acpi_probe_device_table(clksrc); if (!clocksources) pr_crit("%s: no matching clocksources found\n", __func__); } >From the code, we can see that the message was printed because clocksources == 0. This implies that there were no matching nodes on the device tree, and that acpi_probe_device_table returned 0 (which is correct, seeing that the kernel has no ACPI support). This leaves me with two questions: 1 - Is this message a problem indicator? I'm assuming it is, because it's printed with a pr_crit... 2 - Why is for_each_matching_node_and_match not getting any matches? I did not find where __clksrc_of_table is initialized, so maybe it's because it is empty? Cheers, Guilherme

2 2

is there a simple graphic of SATA port multiplier structures?
by Robert P. J. Day April 23, 2019

April 23, 2019

was just asked to take a look at how SATA II port multipliers are supported in the linux kernel and, in the midst of all the code and header files, what seems to be missing is a simple graphic of the main structures and how many of them there are. let me elucidate. imagine, say, a 1-to-4 PMP -- single port multiplexing to four SATA HDDs. from include/linux/libata.h, i can see the ATA-related structures that will be involved, but it would be just ducky if there was a short description as to *how* *many* of each would be involved in this scenario; that is, a simple counting demonstration would be massively informative. first, it appears that there would be a single ata_port structure: struct ata_port { struct Scsi_Host *scsi_host; /* our co-allocated scsi host */ struct ata_port_operations *ops; spinlock_t *lock; /* Flags owned by the EH context. Only EH should touch these once the port is active */ unsigned long flags; /* ATA_FLAG_xxx */ ... snip ... where one of the possible "flags" settings is: ATA_FLAG_PMP = (1 << 19), /* controller supports PMP */ so, having only glanced at all of this so far, i assume i would have one "struct ata_port" with that flag set. on the other hand, it would appear i would have four "struct ata_link"s: struct ata_link { struct ata_port *ap; int pmp; /* port multiplier port # */ ... snip ... finally, what about "struct ata_device": struct ata_device { struct ata_link *link; unsigned int devno; /* 0 or 1 */ ... snip ... not having read any further just yet, it's not clear whether there would be one or four of those, depending on how libata views the storage devices themselves, but i'm sure i'll figure it out as i keep reading. in the end, it would be terrifically useful if there was a simple example of how all these structures connected to support a basic scenario. is that written down anywhere? it doesn't need to be crazy comprehensive, just the enumeration of the structures involved. thoughts? rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

1 0

Need advice about do we need this if condition
by Саша Ковтуненко April 17, 2019

April 17, 2019

hello I am trying to fix https://bugzilla.kernel.org/show_bug.cgi?id=202935 I found cause https://github.com/torvalds/linux/blob/master/fs/cifs/cifsfs.c#L1076 not sure is it correct to remove such condition ? Thanks AK

3 2

Help:get_user_pages_remote - succeed on QEMU, Oops on Android 9 board
by Lev Olshvang April 16, 2019

April 16, 2019

1 1

Is there a "build everything" kernel configuration?
by Jefferson Carpenter April 13, 2019

April 13, 2019

Supposing I want to make a change to the Kernel code and make sure that it all builds (i.e. that I did not cause any compile errors under any configuration), what's best way to do that? Can I just take the default setting for all config options or does there exist a "build everything" mode? - Jefferson Carpenter

2 1