Bug#921193: libmkl-rt: Octave returns wrong results when large arrays are multiplied
Package: libmkl-rt Version: 2019.1.144-4 Severity: important Dear Maintainer, After installing intel-mkl and setting MKL as default BLAS/LAPACK implementation, wrong results are obtained when large arrays are multiplied in octave. Here is the output of a MWE: octave:1> x1=0:10;[x1;x1]*[x1;x1]' ans = 450792059465624 472496191531028 512874660819596 531922410008450 This result is absolutely wrong. All the elements in the 'ans' matrix should have the same value. For example, after changing back to the original BLAS/LAPACK implementation, the result is octave:1> x1=0:10;[x1;x1]*[x1;x1]' ans = 385 385 385 385 which is the right one. Regards Ido -- System Information: Debian Release: buster/sid APT prefers testing APT policy: (500, 'testing'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.19.0-1-amd64 (SMP w/4 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled Versions of packages libmkl-rt depends on: ii debconf [debconf-2.0] 1.5.70 ii libatlas3-base [liblapack.so.3]3.10.3-7+b1 ii libblas3 [libblas.so.3]3.8.0-2 ii libc6 2.28-5 ii libgcc-5-dev 5.4.1-4 ii libgcc-6-dev 6.5.0-1 ii libgcc-7-dev 7.4.0-2 ii libgcc-8-dev 8.2.0-15 ii liblapack3 [liblapack.so.3]3.8.0-2 ii libmkl-locale 2019.1.144-4 ii libmkl-meta-computational 2019.1.144-4 ii libmkl-meta-interface 2019.1.144-4 ii libmkl-meta-threading 2019.1.144-4 ii libomp-7-dev 1:7.0.1-4 ii libomp-dev 1:7.0-47 ii libopenblas-base [liblapack.so.3] 0.3.5+ds-1 libmkl-rt recommends no packages. libmkl-rt suggests no packages. -- debconf information: * libmkl-rt/use-as-default-blas-lapack: true * libmkl-rt/exact-so-3-selections: libblas.so.3, liblapack.so.3, libblas64.so.3, liblapack64.so.3, libmkl-rt/title:
Bug#920730: libmkl-rt: Octave crashes when using MKL as default BLAS/LAPACK implementation
Package: libmkl-rt Version: 2019.1.144-3 Severity: important Dear Maintainer, After installing intel-mkl, and setting MKL as the default BLAS/LAPACK implementation, a severe problem appeared in octave. It crashes on large array product. Here is the output of a MWE which was written in octave's command line: octave:1> x1=0:100; x1*x1' ans = 338350 octave:2> x1=0:1000; x1*x1' ans = 333833500 octave:3> x1=0:1; x1*x1' octave-cli: symbol lookup error: /usr/lib/x86_64-linux- gnu/libmkl_intel_thread.so: undefined symbol: __kmpc_global_thread_num ido@ihlaptop:~ **$ You can see that octave crashed after the 3rd command. This behavior is consistent. It disappears whenever I run: dpkg-reconfigure libmkl-rt and select 'No' when I am asked whether MKL should be the default BLAS/LAPACK implementation. Regards *** Reporter, please consider answering these questions, where appropriate *** * What led up to the situation? * What exactly did you do (or not do) that was effective (or ineffective)? * What was the outcome of this action? * What outcome did you expect instead? *** End of the template - remove these template lines *** -- System Information: Debian Release: buster/sid APT prefers testing APT policy: (500, 'testing'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.19.0-1-amd64 (SMP w/4 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled Versions of packages libmkl-rt depends on: ii debconf [debconf-2.0] 1.5.70 ii libatlas3-base [liblapack.so.3]3.10.3-7+b1 ii libblas3 [libblas.so.3]3.8.0-2 ii libc6 2.28-5 ii libgomp1 8.2.0-14 ii liblapack3 [liblapack.so.3]3.8.0-2 ii libmkl-locale 2019.1.144-3 ii libmkl-meta-computational 2019.1.144-3 ii libmkl-meta-interface 2019.1.144-3 ii libmkl-meta-threading 2019.1.144-3 ii libomp51:7.0-47 ii libomp5-7 1:7.0.1-4 ii libopenblas-base [liblapack.so.3] 0.3.5+ds-1 ii libtbb22018~U6-4 libmkl-rt recommends no packages. libmkl-rt suggests no packages. -- debconf information: * libmkl-rt/exact-so-3-selections: libblas.so.3, liblapack.so.3, libblas64.so.3, liblapack64.so.3, libmkl-rt/title: * libmkl-rt/use-as-default-blas-lapack: true
Bug#731016: Software solution
Jiri brought to my attention that that asus provide firmware upgrade which solved his problem. I found at https://www.asus.com/Optical-Drives-Storage/AN300_External_HDD/HelpDesk _Download/ utilities which correspond my HDD, i.e.: 1. firmware upgrade utility 2. virtual cd removal utility I succeeded to run them under windows (wine didn't helped...). And ... it worked! After the firmware upgrade the block device was generated, though the partition was not recognized. I used the second utility and removed the virtual cd and everything is getting mounted in linux! Ido
Bug#731016: Physical Workaround
Well, I finally found some workaround to that problem. I figured that the problem is with the USB bridge. It appears that the kernel can't deal this bridge. Maybe because it is a USB 3.0 SuperSpeed or it something to do with the ASMEDIA ASM1051 chip. Two LUN are recognized but the HDD LUN is reported to be disconnected (that the meaning of 'PQ:1' in: [...] [ 233.579036] scsi 6:0:0:0: Direct-Access USB3.0 D Super Speed Mass 0PQ: 1 ANSI: 0 [...] I found some 'hardware' work arround to that problem. I bought an other USB HDD enclosure (only 5$ ...) but I made sure it uses USB 2.0. I moved the toshiba HDD into the new enclosure and everything is OK. Even the transfer speed is not so bad. I believe that this bug should be treated in the frame of 'usb_storage' module but this is far beyond my knowledge. If there is way I can contribute to the solution of this bug, let me know. Best Regards Ido Halperin -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#731016: Additional information
I noticed something that might point on the problem. the peripheral qualifier of the HDD logical unit (LUN) is 'PQ:1'. That means that the LUN is currently disconnected (that is according table 46 in http://www.staff.uni-mainz.de/tacke/scsi/SCSI2-08.html). That obviously not true as the HDD works out of the box under windows. For some reason the scsi driver thinks that LUN is disconnected. The question is how windows can see that the LUN is connected and linux cannot? I have more information about this HDD: 1. opened the enclosure and found that it consists of: * Toshiba 2.5" HDD MQ01ABD050 * USB bridge. It is based on ASMEDIA chip - ASM1051. 2. I was able to get some information about the HDD by using the command: **# smartctl -a -d sat /dev/sg2 smartctl 6.2 2013-07-26 r3841 [x86_64-linux-3.12-1-amd64] (local build) Copyright (C) 2002-13, Bruce Allen, Christian Franke, www.smartmontools.org === START OF INFORMATION SECTION === Model Family: Toshiba 2.5" HDD MQ01ABD... Device Model: TOSHIBA MQ01ABD050 Serial Number:722ESCR3S LU WWN Device Id: 5 39 425c803db Firmware Version: AX002J User Capacity:500,107,862,016 bytes [500 GB] Sector Sizes: 512 bytes logical, 4096 bytes physical Rotation Rate:5400 rpm Device is:In smartctl database [for details use: -P show] ATA Version is: ATA8-ACS (minor revision not indicated) SATA Version is: SATA 3.0, 6.0 Gb/s (current: 3.0 Gb/s) Local Time is:Tue Jan 7 14:22:04 2014 IST SMART support is: Available - device has SMART capability. SMART support is: Enabled [...] Ido -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#731016: Can't mount external USB HD Asus AN300
I have just upgraded into kernel 3.12-1-amd64 but the problem persists. Until this issue will be solved I am trying to find some way to manually create the block device for this HDD but I don't know where I should look. I found some pages about adding scsi device and following that I typed: **$ dmesg | tail -n 100 [...] [24690.748752] usb 2-4: new SuperSpeed USB device number 7 using xhci_hcd [24690.765189] usb 2-4: New USB device found, idVendor=0b05, idProduct=17b6 [24690.765196] usb 2-4: New USB device strings: Mfr=2, Product=3, SerialNumber=1 [24690.765200] usb 2-4: Product: Super Speed Mass Storage [24690.765203] usb 2-4: Manufacturer: USB3.0 Device [24690.765205] usb 2-4: SerialNumber: 40201208020A [24690.765975] usb-storage 2-4:1.0: USB Mass Storage device detected [24690.766258] scsi17 : usb-storage 2-4:1.0 [24691.768698] scsi 17:0:0:0: Direct-Access USB3.0 D Super Speed Mass 0PQ: 1 ANSI: 0 [24691.768878] scsi 17:0:0:1: CD-ROMASMT VirtualCDROM PQ: 0 ANSI: 0 [24691.769160] scsi 17:0:0:0: Attached scsi generic sg2 type 0 [24691.769681] sr1: scsi-1 drive [24691.769834] sr 17:0:0:1: Attached scsi CD-ROM sr1 [24691.769920] sr 17:0:0:1: Attached scsi generic sg3 type 5 [24691.820241] xhci_hcd :00:14.0: WARN Event TRB for slot 14 ep 2 with no TDs queued? [24691.960100] ISO 9660 Extensions: Microsoft Joliet Level 3 [24691.960231] ISOFS: changing to secondary root I understand from that output that 17:0:0:0 is my hdd. I tried: **$ echo "0 0 0" > /sys/class/scsi_host/host17/scan but no progress was achieved. Do you know if there is some scsi administration tool that may help me tom manually create the block device file? Ido.
Bug#731016: Can't mount external USB HD Asus AN300
I appologize. I accidently posted the traffic of regular flash-memory disk. This is the traffic for the HDD: USB data USB3.0 Device Super Speed Mass Storage - USB Mass Storage Device ID: 1 Function: IRP_MJ_PNP Minor function: PNP_IRP_UNKNOWN (0x18) Status code: 0xC0BB Status: ERROR ID: 2 Function: IRP_MJ_PNP Minor function: IRP_MN_QUERY_RESOURCE_REQUIREMENTS Status code: 0xC0BB Status: ERROR ID: 3 Function: IRP_MJ_PNP Minor function: IRP_MN_FILTER_RESOURCE_REQUIREMENTS Status code: 0xC0BB Status: ERROR ID: 4 Function: IRP_MJ_PNP Minor function: IRP_MN_START_DEVICE Status code: 0x Status: STATUS_SUCCESS ID: 5 Function: IRP_MJ_PNP Minor function: IRP_MN_QUERY_CAPABILITIES Status code: 0x Status: STATUS_SUCCESS ID: 6 Function: IRP_MJ_PNP Minor function: IRP_MN_QUERY_PNP_DEVICE_STATE Status code: 0x Status: STATUS_SUCCESS ID: 7 Function: IRP_MJ_SYSTEM_CONTROL Status code: 0x Status: STATUS_SUCCESS Ido On Mon, 2013-12-02 at 18:15 +0100, Josua Dietze wrote: [...] > In case there IS a driver required on Windows, you might be able to catch the > initial traffic between this driver and the device once the thing is > re-plugged. In a second step, the initiating command(s) can possibly be > 'replayed' in Linux. This is the usual way usb_modeswitch works. > Note that if a special Windows driver is required for actually ACCESSING the > disk (not just for initialization), then a new Linux driver may be required > as well ... > [...] I don't know if it will help but I tried to capture the USB traffic on Windows station at the time I connected the drive. I got some output but I don't know it's meaning. That what I got by the USB capturing software: USB data SMI Corporation USB DISK - USB Mass Storage Device [...] -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#731016: Can't mount external USB HD Asus AN300
On Mon, 2013-12-02 at 18:15 +0100, Josua Dietze wrote:
[...]
> In case there IS a driver required on Windows, you might be able to catch the
> initial traffic between this driver and the device once the thing is
> re-plugged. In a second step, the initiating command(s) can possibly be
> 'replayed' in Linux. This is the usual way usb_modeswitch works.
> Note that if a special Windows driver is required for actually ACCESSING the
> disk (not just for initialization), then a new Linux driver may be required
> as well ...
>
[...]
I don't know if it will help but I tried to capture the USB traffic on
Windows station at the time I connected the drive. I got some output but
I don't know it's meaning. That what I got by the USB capturing
software:
USB data
SMI Corporation USB DISK - USB Mass Storage Device
ID: 1
Function: IRP_MJ_PNP
Minor function: PNP_IRP_UNKNOWN (0x18)
Status code: 0xC0BB
Status: ERROR
ID: 2
Function: IRP_MJ_PNP
Minor function: IRP_MN_QUERY_RESOURCE_REQUIREMENTS
Status code: 0xC0BB
Status: ERROR
ID: 3
Function: IRP_MJ_PNP
Minor function: IRP_MN_FILTER_RESOURCE_REQUIREMENTS
Status code: 0xC0BB
Status: ERROR
ID: 4
Function: IRP_MJ_PNP
Minor function: IRP_MN_START_DEVICE
Status code: 0x
Status: STATUS_SUCCESS
ID: 5
Function: IRP_MJ_PNP
Minor function: IRP_MN_QUERY_CAPABILITIES
Status code: 0x
Status: STATUS_SUCCESS
ID: 6
Function: IRP_MJ_PNP
Minor function: IRP_MN_QUERY_PNP_DEVICE_STATE
Status code: 0x
Status: STATUS_SUCCESS
ID: 7
Function: IRP_MJ_SYSTEM_CONTROL
Status code: 0x
Status: STATUS_SUCCESS
That's the properties of the device under windows:
USB device details
USB3.0 Device Super Speed Mass Storage - USB Mass Storage Device
File:
\??\USB#VID_0B05&PID_17B6#40201208020A#{a5dcbf10-6530-11d2-901f-00c04fb951ed}
Driver key name: {36fc9e60-c465-11cf-8056-44455354}\0019
Device instance ID: USB\VID_0B05&PID_17B6\40201208020A
Device hardware ID: USB\VID_0B05&PID_17B6&REV_0100
Device service: USBSTOR
Device friendly name:
Device location: Port_#0005.Hub_#0004
Device physical object name: \Device\USBPDO-4
Device descriptor
USB: 0x0210
Device class: 0x00
Device subclass: 0x00
Device protocol: 0x00
Max packet size: 0x40 (64)
Vendor: 0x0B05 (ASUSTek Computer, Inc.)
Product ID: 0x17B6
Device ID: 0x0100
Manufacturer: 0x02 (USB3.0 Device)
Product: 0x03 (Super Speed Mass Storage)
Serial number: 0x01 (40201208020A)
Number of configurations: 1
Connection status: Device connected
Current config value: 0x01
Device bus speed: High (USB 2.0, 480 Mbit/s)
Device address: 0x02
Open pipes: 2
Pipe information
Endpoint address: 0x81, Input
Transfer type: Bulk
Max packet size: 0x0200 (512)
Interval: 0x00 (0)
Pipe information
Endpoint address: 0x02, Output
Transfer type: Bulk
Max packet size: 0x0200 (512)
Interval: 0x00 (0)
Configuration descriptor
Total length: 0x0020
Number of interfaces: 1
Configuration value: 0x01
Configuration ID: 0x00
Attributes: 0xC0 (Bus powered, Self powered)
Max power: 0x00 (0 Ma)
Interface descriptor
Interface number: 0x00
Alternate setting: 0x00
Num of endpoints: 2
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#731016: Can't mount external USB HD Asus AN300
That what I got: **$ sudo lsusb -v -d 0b05:17b6 Bus 002 Device 004: ID 0b05:17b6 ASUSTek Computer, Inc. Device Descriptor: bLength18 bDescriptorType 1 bcdUSB 3.00 bDeviceClass0 (Defined at Interface level) bDeviceSubClass 0 bDeviceProtocol 0 bMaxPacketSize0 9 idVendor 0x0b05 ASUSTek Computer, Inc. idProduct 0x17b6 bcdDevice1.00 iManufacturer 2 USB3.0 Device iProduct3 Super Speed Mass Storage iSerial 1 40201208020A bNumConfigurations 1 Configuration Descriptor: bLength 9 bDescriptorType 2 wTotalLength 44 bNumInterfaces 1 bConfigurationValue 1 iConfiguration 0 bmAttributes 0xc0 Self Powered MaxPower0mA Interface Descriptor: bLength 9 bDescriptorType 4 bInterfaceNumber0 bAlternateSetting 0 bNumEndpoints 2 bInterfaceClass 8 Mass Storage bInterfaceSubClass 6 SCSI bInterfaceProtocol 80 Bulk-Only iInterface 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x81 EP 1 IN bmAttributes2 Transfer TypeBulk Synch Type None Usage Type Data wMaxPacketSize 0x0400 1x 1024 bytes bInterval 0 bMaxBurst 15 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x02 EP 2 OUT bmAttributes2 Transfer TypeBulk Synch Type None Usage Type Data wMaxPacketSize 0x0400 1x 1024 bytes bInterval 0 bMaxBurst 15 Binary Object Store Descriptor: bLength 5 bDescriptorType15 wTotalLength 22 bNumDeviceCaps 2 USB 2.0 Extension Device Capability: bLength 7 bDescriptorType16 bDevCapabilityType 2 bmAttributes 0x0002 Link Power Management (LPM) Supported SuperSpeed USB Device Capability: bLength10 bDescriptorType16 bDevCapabilityType 3 bmAttributes 0x00 wSpeedsSupported 0x000e Device can operate at Full Speed (12Mbps) Device can operate at High Speed (480Mbps) Device can operate at SuperSpeed (5Gbps) bFunctionalitySupport 1 Lowest fully-functional device speed is Full Speed (12Mbps) bU1DevExitLat 10 micro seconds bU2DevExitLat2047 micro seconds Device Status: 0x000d Self Powered U1 Enabled U2 Enabled Regarding the line wrapping, it messed out in the copy-paste procedure. I am sorry for that ... It happend not only for Asus. It seems that this bug appeared 3 years ago for external Corsair disk but no solution was found: http://superuser.com/questions/225314/cant-mount-linux-usb-disk-it-just-create-dev-sg-device-but-no-dev-sd http://forum.corsair.com/v3/showthread.php?t=91434 I also wanted to say to you guy's that I very appreciate your efforts in solving this issue. Thank you. Ido -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#731016: Can't mount external USB HD Asus AN300
Ok. I already tried it on Windows and it worked but I don't remember if Windows asked a permission to install a driver. I will check it again and tell you what happens. Though, it might take me a little time. Ido \On Mon, 2013-12-02 at 18:15 +0100, Josua Dietze wrote: > usb_modeswitch can only help if the device needs a special initialization > before becoming useable in a standard way. > > The indication for this is that Windows will ask for driver installation when > the disk is plugged in for the first time. > > If it works on Windows right away with no installation required at all, then > the USB storage driver of Linux is the place to try fixes or work-arounds. > > Googling for this device brought up only cries for help w/ respect to Linux - > so far, I could not find any indication of a successful connection. The > official product page lists only OS X and Windows as supported systems. > > In case there IS a driver required on Windows, you might be able to catch the > initial traffic between this driver and the device once the thing is > re-plugged. In a second step, the initiating command(s) can possibly be > 'replayed' in Linux. This is the usual way usb_modeswitch works. > Note that if a special Windows driver is required for actually ACCESSING the > disk (not just for initialization), then a new Linux driver may be required > as well ... > > Josua Dietze -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#731016: closed by Ben Hutchings (Re: Bug#731016: can't mount external USB HDD - /dev/sdb wasn't created (but /dev/sg2 does...))
I installed the package 'usb-mode-switch' but the problem persists. It's not solved yet. On Sun, 2013-12-01 at 01:12 +, Debian Bug Tracking System wrote: > This is an automatic notification regarding your Bug report > which was filed against the linux-image-3.11-2-amd64 package: > > #731016: can't mount external USB HDD - /dev/sdb wasn't created (but /dev/sg2 > does...) > > It has been closed by Ben Hutchings . > > Their explanation is attached below along with your original report. > If this explanation is unsatisfactory and you have not received a > better one in a separate message then please contact Ben Hutchings > by > replying to this email. > > -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#730088: udev won't creat block device for external usb hdd
After looking at the output of 'udevadm' I noticed that the kernel does not generates block device file for the HDD. In that case I assume that this might be kernel bug and not udev's bug. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#619799: Evolution and CTL text
I confirm this bug. When composing mail with hebrew letters and latin/number letters in the same row, pressing 'Home' or 'End' make evolution to freeze. Ido Halperin -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
